perspective

A “good death”: perspectives of Muslim

patients and health care providers
Mohamad A. Tayeb, Ersan Al- Zamel, Muhammed M. Fareed, Hesham A. Abouellail
From the North West Armed Forces Hospitals, Tabuk, Saudi Arabia

Correspondence: Dr. Mohamad Abdurrahman Tayeb · Hematology & Oncology Center, North West Armed Forces Hospitals, PO Box 100,
Tabuk, Saudi Arabia · drtayebweb@hotmail.com · Submitted: December 2009 Accepted: March 2010

Ann Saudi Med 2010; 30(3): 215-221

PMID: **** DOI: 10.4103/0256-4947.62836

BACKGROUND AND OBJECTIVES: Twelve “good death” principles have been identified that apply to
Westerners. This study aimed to review the TFHCOP good death perception to determine its validity for Muslim
patients and health care providers, and to identify and describe other components of the Muslim good death
perspective.
SUBJECTS AND METHODS: Participants included 284 Muslims of both genders with different nationalities and
careers. We used a 12-question questionnaire based on the 12 principles of the TFHCOP good death definition,
followed by face-to-face interviews. We used descriptive statistics to analyze questionnaire responses. However,
for new themes, we used a grounded theory approach with a “constant comparisons” method.
RESULT: On average, each participant agreed on eight principles of the questionnaire. Dignity, privacy, spiri-
tual and emotional support, access to hospice care, ability to issue advance directives, and to have time to say
goodbye were the top priorities. Participants identified three main domains. The first domain was related to faith
and belief. The second domain included some principles related to self-esteem and person>s image to friends
and family. The third domain was related to satisfaction about family security after the death of the patient.
Professional role distinctions were more pronounced than were gender or nationality differences.
CONCLUSION: Several aspects of «good death,» as perceived by Western communities, are not recognized as
being important by many Muslim patients and health care providers. Furthermore, our study introduced three
novel components of good death in Muslim society.

T
he basic duty of health care providers is to
preserve human health and life, but we should
realize that death is the inevitable destiny of
mankind predetermined by Allah (the creator of the
universe). In cases in which recovery is hopeless, the role
of health care providers does not end but rather is modi-
fied.1 Health care providers need to help patients and
families minimize suffering and maximize comfort by
offering appropriate medical care that is neither exces-
sive nor negligent. This principle cannot be applied in
10% to 20% of cases in which death happens suddenly
(myocardial infarction, arrhythmia, intracranial bleed-
ing, aortic dissection, accident), but in 80% to 90% of
cases, in which death is predicted,2 both the patient and
the medical team are aware of its imminence. This is
known as anticipated death, which must be planned if
all goes well.
The whole subject of death and dying has been ad-
dressed in the West by pioneers such as Elisabeth
Kubler-Ross and Raymond Moody: “Looking deeply
into the way that we care for the dying person, they have
shown that with unconditional love, and a more enlight-
ened attitude, death can be a peaceful, even transforma-
tive experience.”3 However, dealing with a patient at the
end of life, aiming to improve quality of life, and plan-
ning for end-of-life care requires a good understanding
of that patient’s state from a medical, social, spiritual,
and psychological point of view. To date, this important
topic is understudied in schools of medicine, and the ex-
tensive literature on medical errors shows that medical
experts are rarely blamed for “bad” deaths.4 The authors
of the final report on The Future of Health and Care of
Older People (TFHCOP) have identified 12 principles
of a good death (Table 1).5,6
Most of the conducted studies examined the concept
of good death in elderly whites, but no comparable qual-
itative data exist regarding other races.7 No qualitative
study has been conducted in a Muslim society, and pub-

Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net 215

 perspective
Table 1. Principles of a end-of-life care according to the report,
The Future of Health and Care of Older People.
1. To know when death is coming, and to understand what
can be expected.
2. To be able to retain control of what happens.
3. To be afforded dignity and privacy.
4. To have control over pain relief and other symptom control.
5. To have choice and control over where death occurs (at
home or elsewhere).
6. To have access to information and expertise of whatever
kind is necessary.
7. To have access to any spiritual or emotional support
required.
8. To have access to hospice care in any location, not only in
hospital.
9. To have control over who is present at the time when the
end comes.
10. To be able to issue advance directives, which ensure
wishes are respected.
11. To have time to say goodbye, and control over other
aspects of timing.
12. To be able to leave when it is time to go, and not to have
life prolonged pointlessly.
lished papers on this important topic have focused main-
ly on reviewing Islamic teachings and the literature.8
The purpose of this study was to review the
TFHCOP perception of good death to determine its va-
lidity for Muslim patients and health care providers and
to identify and describe other elements for a good death.
We tried to elaborate on the concept of good death in
our society. The Western definition of “good death” was
chosen rather than the Ghanaian or Tibetan one since
our medical practice depends mainly on Western refer-
ences; however, there should be some degree of flexibility
in applying the Western perspective since death beliefs
have religious and cultural backgrounds.
SUBJECTS AND METHODS
Participants were recruited by a random sampling tech-
nique per category from King Abdul-Aziz Military
Hospital and from a rehabilitation center and long-stay
nursing unit in King Khalid Military Hospital in Tabuk.
Since most people die surrounded by medical profes-
sionals, even when dying occurs at home, a full spectrum
of persons involved with end-of-life care-physicians,
nurses, social workers, religious officers, patients with
life-limiting diseases, and their home caregivers (fam-
ily members) were included in the survey.12,13 Patients
and family members were recruited mainly from the
216
A GOOD DEATH
Hematology and Oncology Center, especially those with
advanced malignancy or those who were under palliative
care. Death is a multi-dimensional process that involves
physical symptom management as well as certain reli-
gious, social, and psychological needs of the dying and
their loved ones.14 Non-physical issues may differ widely,
which can be attributed mainly to religious and cultural
differences. Therefore, we expected that Muslim partici-
pants may perceive these issues in a completely different
manner. Approval of the study was obtained from the
local research ethics committee.
Each participant was interviewed as well as adminis-
tered a questionnaire. Before they were administered the
questionnaire, they were made privy to the nature of the
research and its purpose by asking the following ques-
tions: 1) What constitutes a good death and a bad one?
2) Can we develop our own definitions in keeping with
our religious principles and cultural values? 3) Is this 12-
principle definition valid as a definition of a good death in
Islamic culture? 4) Please express your opinions freely. Do
not be influenced by others’ opinions. 5) Share your own
experiences and feelings, and not what you have heard
from others. 6) Express your opinions. There is no wrong
or right answer. 7) We would like to know your opinions.
We developed a questionnaire on the basis of the
TFHCOP’s “12 principles of a good death.” The ques-
tionnaire comprised 12 closed-ended questions to which
the participants had to answer in “Yes” or “No.” The ques-
tionnaires were provided in Arabic or English, depend-
ing on the participants’ preferences (Appendices 1, 2). To
ensure the validity of the questionnaire, before conduct-
ing the main interviews, we analyzed the questionnaire
in both languages with 10 participants of different sub-
groups. We ensured that the questions were simple and
easy to understand, so that the respondents did not have
any difficulty in interpreting and answering the questions.
Some participants, however, had difficulty in choosing a
reply to a few questions. In such cases, we clarified the
questions and explained each option by giving examples.
We then let them determine how essential they found the
point and to choose their answer accordingly.
We used a qualitative approach to draw out previ-
ously unexplored aspects and components of the Muslim
perspective. We conducted in-depth, open-ended, and
face-to-face interviews and content analysis. We did not
impose any theoretical assumptions a priori and instead
let the participants suggest any principles that they con-
sidered to be essential.
After completing the questionnaire, the participants
were asked to express and write their opinions on the
principles (not included in the previous 12 points) that
constitute a good death. We convened focus groups, each
Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net
A GOOD DEATH perspective
Table 2. Number and frequencies of responses to each principle by group.

Career
Sex Nationality
Total Health care Health care Patients
providers providers
Males Females Saudi Non-Saudi Group 1 Group 2

284 164 120 147 137 181 77 26

Principle 1: Timing of death 80 (28%) 48 (29%) 32 (27%) 41 (28%) 39 (28%) 50 (28%) 24 (31%) 26 (23%)
Principle 2: Control of what
177 (62%) 99 (60%) 78 (65%) 90 (61%) 87(64%) 115(64%) 47(61%) 16 (62%)
happens
Principle 3: Dignity and privacy 271 (95%) 156 (95%) 115 (96%) 144 (98%) 127 (93%) 176 (97%) 71 (92%) 24 (92%)
Principle 4: Pain and other
246 (87%) 141 (86%) 105 (88%) 132 (90%) 114 (83%) 163 (90%) 60 (78%) 23 (88%)
symptom relief
Principle 5: Where death
127 (45%) 70 (43%) 57 (48%) 70 (47%) 57 (41%) 86 (48%) 30 (39%) 11 (42%)
occurs
Principle 6: Access to
116 (41%) 50 (30%)* 66 (55%)* 63 (43%) 53 (39%) 82 (45%)* 21 (27%)* 13 (50%)*
necessary information
Principle 7: Spiritual or
255 (90%) 146 (89%) 109 (91%) 135 (92%) 120 (88%) 166 (92%) 66 (86%) 23 (88%)
emotional support
Principle 8: Access to hospice
255 (90%) 144 (88%) 111 (93%) 135 (92%) 120 (88%) 165 (91%) 66 (86%) 24 (92%)
care
Principle 9: Control over who is
108 (38%) 64 (39%) 44 (37%) 59 (40%) 49 (36%) 71 (39%) 30 (39%) 7 (27%)
present
Principle 10: To issue advance
240 (85%) 139 (85%) 101 (84%) 126 (40%) 114 (83%) 155 (86%) 63 (82%) 22 (85%)
directives
Principle 11: To say goodbye 209 (74%) 119 (73%) 90 (75%) 108 (73%) 99 (72%) 133 (73%) 57 (74%) 19 (73%)
Principle 12: To leave when it is
207 (73%) 110 (75%) 97 (71%) 109 (74%) 98 (72%) 137 (76%) 50 (65%) 20 (77%)
time to go

comprising two to seven participants, over a 2-month pe-
riod. The interviews usually lasted for around 30 minutes
per group.
Univariate descriptive statistics (frequencies, percent-
ages) were used to analyze responses to each point of the
questionnaire. Percentages are reported as whole num-
bers. For bivariate analysis, differences in participant char-
acteristics among different questions were determined by
using the chi-square test. P values of <.05 were consid-
ered statistically significant. For new themes we followed
a grounded theory approach with a “constant compari-
sons” method and its related open and axial coding tech-
niques.9,10 Trustworthiness was noted when participants
respond affirmatively to researchers’ interpretations.11
RESULTS
Researchers interviewed 284 participants. Focus group
participants ranged in age from 18 to 71 years (mean age,
37 years). The study sample consisted of 164 males (58%)
and 120 females (42%). The study included participants
of several different nationalities: 147 Saudi (52%), 42
Egyptian (15%), 23 Syrian (8%), 17 Pakistani (6%), 12
Sudanese (4%), 11 Filipino (4%), 11 Jordanian (4%), 10
Malaysian (4%), 7 Indian (3%), 2 South African (1%),
1 Moroccan (<1%), and 1 Nigerian (<1%). The most
common groups among the participants were 93 nurses
(33%), 88 physicians (31%), 43 patients’ relatives (15%),
26 patients (9%), 16 social workers (6%), 14 clinical
pharmacists (5%), and 4 hospital religious officers (1%).
Some focus group members were concerned about our
society’s tendency to consider death as taboo, something
that human beings cannot interfere with. Participants in-
formed us of cases in which health care providers avoided
end-of-life discussions because they believed that it is be-
yond our control as humans.
On average, each participant agreed with the eight

Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net 217

 perspective
Percentage
Principles
Figure 1. Percentage of responses to each principle by group.
Table 3. Principles of a good death from the Muslim perspective.
· Aspects related to faith and relationship with Allah
· Aspects related to self-esteem and person’s image in the eyes of relatives
· Aspects related to concerns about family security
· To be afforded dignity and privacy
· To have access to any spiritual or emotional support required
· To have access to hospice care in any location, not only in hospital
· To be able to have control over pain relief and other symptom control
· To be able to issue advance directives, which ensure wishes are respected
· To have time to say goodbye and control over other aspects of timing
· To be able to leave when it is time to go, and not to have life prolonged pointlessly
· To be able to retain control of what happens
principles of the questionnaire, so we selected the top
eight as the important principles for our society. Table 2
shows the frequencies and percentage of “yes” answers for
each principle according to nationality, gender, and career.
Health care providers were categorized into two groups:
Group 1 included health care providers who are involved
mainly in medical issues, i.e. physicians and nurses;
Group 2 included the other members of the multi-dis-
ciplinary care team who are involved mainly in other as-
pects of care, i.e. patients’ relatives, social workers, clinical
pharmacists, and hospital religious officers.
Participants identified certain aspects of quality of
death that are not mentioned in the Western literature
218
A GOOD DEATH
but prove to be essential for Muslims. These can be sum-
marized in three main domains:
Religious faith and beliefs
The first domain is related to a Muslim’s faith, belief, and
preferences during the dying process. This includes mat-
ters like 1) being sure that somebody is there to prompt
him with Shahadah (bearing witness that there is no true
God but Allah and Muhammad is verily His Servant
and His Messenger) as a final statement of faith, 2) the
presence of someone at the bedside to recite chapters
of the Noble Qur’an, 3) to die in a position facing the
holy mosque in Makkah, 4) to die in a holy place (e.g.,
Madinah, Makkah, or mosque) or in a holy time (e.g., in
Ramadan or on a Friday).
Self-esteem and body image
The second domain includes some principles related to
the patient’s self-esteem and image in his friends’ and
relatives’ eyes, by avoiding post-mortem distortions,
deformities, septic wounds, or bad odors by maintain-
ing continence and keeping the body and clothes free of
urine, stool, and vomit, and making sure the body has a
normal appearance after death.
Concerns about family security
The third domain is related to the patient’s satisfaction
about his relatives. In other words, he needs to feel that
his family will be secure and have no trouble after his
death, so he will not be worried about them. This relates
primarily to economic and social concerns.
It is worth mentioning that data obtained from a
sub-group of Saudis reflected the same eight principles
selected by major subgroups including females, health
care professionals (Group 1), non-medical health care
providers (Group 2), patients’ home caregivers, and pa-
tients themselves (Figure 1). However, the order of se-
quence of these selected eight points was slightly differ-
ent. The resultant responses were almost identical to all
questions with minimal differences among groups. On
the other hand, the three new themes reflect the com-
mon ground shared by participants. However, we found
remarkable differences among groups. In fact, profes-
sional role distinctions were more pronounced than
were gender or nationality differences. For example,
most physicians and nurses were highly attuned to the
needs of good hygiene and preservation of a patient up
until his last breath. Religious officers discussed faith
and spiritual support and were the only group to dis-
cuss postmortem preservation of configuration. Social
workers’ discussions were more psychosocial in nature
and focused on being reassured about the family’s future
Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net
A GOOD DEATH
after their death. All themes were present in the patient
focus group, but this group showed more concern about
faith and their relationship with Allah Almighty.
DISCUSSION
This study reveals that several aspects of “good death” as
perceived by Western communities are not recognized
as being of special importance by many Muslim patients
and health care providers. Furthermore, our study intro-
duced three novel components of good death in Muslim
society. The definition of “good death” that is acceptable
to Muslim patients and health care providers consists
of 11 principles (Table 3). There was some kind of con-
sensus on the importance of dignity and privacy, which
can be explained by the Islamic perspective that respects
human dignity and privacy and regards each as a funda-
mental pillar of Shari’a (Islamic Law). Secondly, Muslims
by and large value spiritual and emotional support. They
also believe that death is closely linked to faith. Therefore,
most participants appreciated the importance of access to
any needed spiritual or emotional support.
Control over pain and other symptoms is generally
required by most people.14 Muslims perceive suffering
as atonement for one’s sins. This interpretation helps pa-
tients and family members cope with diseases. However,
it does not belittle the fact that every effort should be
made to relieve suffering. In our study, we met partici-
pants who prefer not to receive pain management be-
cause they believe that if a patient feels pain and shows
patience, he will be rewarded more by Allah and will be
more pure. We have to respect this wish because patients
are granted the full right to accept or to refuse medical
intervention despite the fact that Islamism encourages
seeking treatment and there is a lack of religious evidence
to support this point of view.8 Principle 4 may be modi-
fied to suit Muslim patients: “To have control over pain
relief and other symptom control according to patient
desire or preference.”
Further, while conducting interviews, we realized that
some medical staff, including physicians, are not aware
of “advance directives.” However, most of the participants
preferred to issue advance directives, i.e. the right to re-
fuse any therapeutic intervention, after the concept was
explained to them. This study emphasizes that advance
directives are underused in the medical practice of this
hospital and that we have to encourage adopting it in
our hospitals, especially since this study revealed that it
is widely accepted by Muslim participants. It is very com-
mon in Arabic and Islamic societies to contact relatives
and loved ones who may want to see the patient before
death, but a physician has to suggest this when he feels
that death is imminent and the patient is about to take
Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net
perspective
his last breath. This is also standard practice in the West:
staff members invite relatives to be present when death is
impending to quietly say their goodbyes.15 If the number
of visitors exceeds the available space, health care provid-
ers are expected to express a high degree of sensitivity if it
becomes necessary to intervene in such situations.8
One cause of increased suffering is facing a lonely
death in hospitals surrounded by strangers and advanced
technologies and procedures that needlessly prolong the
dying process without realizing the fact that the end of
life is approaching.16,17 This may be a good reason why
most participants chose to be able to leave when it is time
to die and not to have life prolonged pointlessly. More
importantly, Muslims do not generally believe that life
is pointless, even when it is associated with a significant
amount of suffering. They believe that Allah has the ul-
timate wisdom and is the most merciful and He will re-
ward those who express patience and satisfaction when
inflicted with disease or suffering. This, on the other
hand, emphasizes the importance of being honest with
patients, telling them about their prognosis, and giving
them detailed explanations of Do Not Resuscitate or-
ders. This may help in reassuring patients and makes
them feel more comfortable that they will not receive any
futile intervention.
Muslims believe that it is impossible to tell the time or
place of death by any means and that it is only known by
Allah Almighty who predetermines the exact timing and
place of death. This may explain why the agreement rate
for these principles was the least in this study, although
many participants confirmed that it would be nice to
have a “death alarm” that would enable the person to re-
pent for his sins and ask for forgiveness. This supports
the view that the patient and his family prefer to be given
a less definitive answer. Surprisingly, if given a choice of a
specific place to die, some participants repeatedly stated
that they preferred to die in a holy place like a mosque, or
in Makkah or Madinah. In this case, health care provid-
ers have to respect the patient’s wish and may sometimes
even be asked by the patient to help in persuading his/her
family to grant this wish. This may be achieved by build-
ing a good rapport with the family and being honest and
open with them regarding the patient’s status, prognosis,
and goals of care at that stage. This approach may help to
decrease the patient’s stress and prevent family members
from feeling guilty about leaving him to die outside the
hospital. Readers may note that we talk about the patient
and his family as one unit, which contradicts Western
practice and the concept of medical confidentiality. This
could be attributed to the distinct nature of the Muslim
family and its composite interrelations and strong ties.
This study has some limitations related to method and
219
 perspective
study design. A “Yes/No” question format may not be the
best way to measure participants’ preferences in this type
of research. A five-point Likert scale would have been a
better assessment tool because it would allow the partici-
pants to choose from a range of options that reflect the
degree to which they agree or disagree with a statement
and also to abstain from committing themselves to one
response or the other if they were not sure. Although the
fact that we obtained similar responses from all subgroups
may support test reliability, further studies are required to
assess the validity of using a questionnaire for this kind of
topic. The questionnaire was provided in one of two lan-
guages, which may bear the potential for inter-rater vari-
ability and language translation inaccuracy. Researchers
conducted a pilot study with 10 participants who received
the questionnaire in one language and then in the other
language. Proper modifications were made to ensure re-
ceipt of the same response. There was a preliminary as-
sumption that principles of good death among whites,
as reported by a TFHCOP, are applicable to a group of
Muslim participants. This was obvious by adopting these
principles as leading questions addressed to Muslim par-
ticipants. The authors suggest that similar studies can be
repeated using modified principles that are in better con-
cordance with Islamic culture. Although our participants
represented a broad range of ages, nationalities, careers,
and educational levels, all were recruited from Northwest
Armed Forces Hospitals; therefore, our findings may not
reflect the perceptions in other areas. A large study on a
national or international basis may be of importance to
address this issue.
Our study has implications for both education and
practice. Health care providers must understand their
patients’ preferences and respect their needs regardless
of what the health care provider believes. Physicians and
nurses should do their best to care for each patient’s ap-
pearance and hygiene, as this is of great importance to
Muslims. This can be achieved by keeping the clothes and
body of the patient free from urine, stool, vomit, or blood
220
A GOOD DEATH
whenever possible. They should be ready to help the pa-
tient to take a bath more frequently as well. To reduce
postmortem disfigurement, jaw fixation and eye closing
must be done immediately at the time of death. All pa-
tients should be reassured that this is the practice with all
deceased Muslims in the hospital. The rites of washing,
shrouding, funeral prayers, and burial should follow as
soon as possible. In addition to expressing empathy, the
major role of the health care team at this stage is timely
documentation to prevent any unnecessary delay in pro-
ceeding with funeral rites. Physicians should be reminded
that they are not alone when caring for dying patients;
many other health care providers (nurses, social workers,
and religious officers) are available to provide comprehen-
sive care. Physicians may ask a screening question regard-
ing any faith-related concerns and then ask whether the
patient or relatives would like to speak in greater depth
with a religious officer (a Sheikh or Imam). In addition,
the physician should ask the patient a screening ques-
tion to figure out any concerns related to his relatives
and make sure that he has no other worries. The physi-
cian can then consult the social worker to help him in this
respect. However, the palliative care team may need to
identify families with suboptimal resources, provide the
necessary support during bereavement, and reassure the
patient that his family will receive economic support after
his death. This is not the physician’s primary duty, yet he
should explore this point and ask for proper consultation,
if necessary.
It is important that the medical curriculum covers this
issue in detail. The Islamic perspective of “good death” must
be included in health care services, professional codes, and
care plans or missions for end-of-life care organizations
and institutions in Islamic countries. This is more compli-
cated than simply translating Western literature and try-
ing to apply it to our practice that exists in a society with
different cultural and religious backgrounds; care must be
taken to develop such information in a Muslim-centered
manner.
Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net
A GOOD DEATH
REFERENCES
1. Aljarallah JS, Abu Aisha H, Al-Mazroo AM,
Al-wazzan KA, Al-Zahrani SA Ethics of the Medi-
cal Profession. 2 nd ed. Riyadh: Saudi Council of
Medical Specialities; 2003.
2. Wellens J, Gorgels AP, de Munter H. Sud-
den death in the community. Electrophysiology
2003;14:S104-7.
3. Rinpoche S. The Tibetan Book of living and dy-
ing. UK: Oxford Press; 1999.
4. Kohn LT, Corrigan JM, Donaldson MS, eds.
To err is human: Building a safer health system.
Washington, DC: National Academy Press; 2000.
5. Debate of the Age Health and Care Study
Group. The future of health and care of older
people: the best is yet to come. London: Age
Concern; 1999.
6. Smith R. A good death: An important aim for
health services and for us all. BMJ 2000;320:129-
30.
7. Vig EK, Pearlman RA. Good and bad dying
from the perspective of terminally ill men. Arch
Intern Med 2004;164:977-81.
8. al-Shahri MZ, al-Khenaizan A. Palliative care
for Muslim patients. J Support Oncol 2005;3:432-
6.
9. Singer PA, Martin DK, Kelner M. Quality
end-of-life care: Patients’ perspectives. JAMA
1999;281:163-8.
10. Corbin J, Strauss A. Basics of qualitative re-
search. Thousand Oaks. CA: Sage; 1990.
11. Krueger R. Focus groups: A practical guide
for applied research. Thousand Oaks. 2 nd ed.
CA: Sage; 1994.
12. Saunders C. The management of terminal
perspective
disease. London: Arnold; 1978.
13. Steinhauser KE, Clipp EC, McNeilly M, Chris-
takis NA, McIntyre LM, Tulsky JA. In search of
a good death: Observations of patients, families,
and providers. Ann Intern Med 2000;132:825-32.
14. Hanson LC, Danis M, Garrett J. What is wrong
with end-of-life care? Opinions of bereaved family
members. J Am Geriatr Soc 1997;45:1339-44.
15. Tsai E. Should family members be present
during cardiopulmonary resuscitation? N Engl J
Med 2002;346:1019-21.
16. Muller JH. Shades of blue: The negotiation of
limited codes by medical residents. Soc Sci Med
1992;34:885-98.
17. Harvey J. The technological regulation of
death: With reference to the technological regu-
lation of birth. Sociology 1997;31:4,7.

Appendix 1. Questionnaire in Arabic. Appendix 2. Questionnaire in English: What do you think about
these principles? Do you think they are well-matched to your
conception of a “good death?”

Ann Saudi Med 30(3) May-June 2010 www.saudiannals.net 221

 Medicine
OBSERVATIONAL STUDY
A Cross-Cultural Study on Behaviors When Death Is
Approaching in East Asian Countries
What Are the Physician-Perceived Common Beliefs and Practices?
Shao-Yi Cheng, MD, MSc, DrPH, Sang-Yeon Suh, MD, MPH, PhD, Tatsuya Morita, MD,
Yasuhiro Oyama, PhD, Tai-Yuan Chiu, MD, MHS, Su Jin Koh, MD, PhD, Hyun Sook Kim, PhD, RN, MSW,
Shinn-Jang Hwang, MD, Taeko Yoshie, MA, and Satoru Tsuneto, MD, PhD
Abstract: The primary aim of this study was to explore common
beliefs and practices when death is approaching in East-Asian countries.
A cross-sectional survey was performed involving palliative care
physicians in Japan, Korea, and Taiwan. Measurement outcomes were
physician-perceived frequencies of the following when patient death
was approaching: (1) reluctance to take part in end-of-life discussions,
(2) role of family members, (3) home death, and (4) circumstances
surrounding death.
A total of 505, 211, and 207 responses were obtained from Japanese,
Korea, and Taiwan physicians, respectively. While 50% of the Japanese
physicians reported that they often or very often experienced families as
being reluctant to discuss end-of-life issues, the corresponding figures
were 59% in Korea and 70% in Taiwan. Two specific reasons to avoid
end-of-life discussion, ‘‘bad things happen after you say them out loud’’
and ‘‘a bad life is better than a good death’’ were significantly more
frequently observed in Taiwan. Prioritizing the oldest of the family in
breaking bad news and having all family members present at the time of
death were significantly more frequently observed in Korea and Taiwan.
Half of Taiwanese physicians reported they often or very often experi-
enced the patients/family wanted to go back home to die because the
soul would not be able to return from the hospital. In all countries, more
than 70% of the physicians reported certain family members were
Editor: Wen-Hung Wang.
Received: June 25, 2015; revised: July 30, 2015; accepted: August 19,
2015.
From the Department of Family Medicine, College of Medicine and
Hospital, National Taiwan University, Taipei, Taiwan (SYC), Department
of Family Medicine, Dongguk University Ilsan Hospital, Dongguk
University School of Medicine, Seoul, South Korea (S-YS), Department
of Palliative and Supportive Care, Palliative Care Team, and Seirei
Hospice, Seirei Mikatahara General Hospital, Japan (TM), Division of
Clinical Psychology, Kyoto University, Kyoto, Japan (YO), Department of
Family Medicine, College of Medicine and Hospital, National Taiwan
University, Taipei, Taiwan (T-YC), Department of Hematology and
Oncology, Ulsan University Hospital, University of Ulsan College of
Medicine, Ulsan, South Korea (SJK), Department of Social Welfare, Korea
National University of Transportation, Chungju City, South Korea (HSK),
Department of Family Medicine, Taipei Veterans General Hospital and
National Yang Ming University, School of Medicine, Taipei, Taiwan (S-
JH), Graduate School of Education, Kyoto University, Kyoto, Japan (TY),
Department of Multidisciplinary Cancer Treatment, Graduate School of
Medicine, Kyoto University, Kyoto, Japan (ST).
Correspondence: Shao-Yi Cheng, Department of Family Medicine, College
of Medicine and Hospital, National Taiwan University, 7, Chung Shan
S. Rd, Taipei, Taiwan (e-mail: scheng2140@gmail.com).
The authors report no conflicts of interest to disclose.
Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved.
This is an open access article distributed under the Creative Commons
Attribution-NonCommercial-NoDerivatives License 4.0, where it is
permissible to download, share and reproduce the work in any medium,
provided it is properly cited. The work cannot be changed in any way or
used commercially.
ISSN: 0025-7974
DOI: 10.1097/MD.0000000000001573
Medicine
Volume 94, Number 39, October 2015
®
expected to care for the patient at home. At the time of death, while
no Japanese physicians stated that they often experienced patients
wanted a religious person to visit, the corresponding figure in Korean
and Taiwan was about 40%. Uncovered expression of emotion was
significantly frequently observed in Korean and Taiwan, and 42% of the
Japanese physicians reported family members cleaned the dead body of
the patient themselves.
There seem to be significant intercountry differences in beliefs and
practices when death is approaching in East Asian countries. Future
studies on direct observations of patients and families are needed.
(Medicine 94(39):e1573)
Abbreviation: ANOVA = analysis of variance.
INTRODUCTION
A n understanding of cultural differences is very important for
providing patient- and family-centered end-of-life care.
What patients and families believe is usual or appropriate when
death is approaching is heavily influenced by culture, and to
provide appropriate care for imminently dying patients, an under-
standing of the culture of the patients and family members is
essential.1–4 Nonetheless, to the best of our knowledge, there
have been no large-scale studies regarding what is usual when
death is approaching, except for review articles based on the
experience of clinicians. Existing cross-cultural studies have been
focused on patient autonomy, information disclosure, communi-
cation, decision-making, hydration, pain management, and with-
drawal/withholding of life sustaining treatments.5–12
East Asia is traditionally regarded as a typical family-
centered region that is based on Confucian culture.4 Japan,
Korea, and Taiwan, however, have considerable differences in
cultural traditions, and each country could have unique
traditions about death and dying. For example, several studies
indicated that, in Taiwan, families believe that the soul will not
be able to return from the hospital if patients die in hospitals;
this could have a great deal of influence on the patient-pre-
ferred place of death, that is, many Taiwanese want to go back
home to die even if they are unconscious.4,13 – 15 In Korea, it is
suggested that the concept of filial piety (devotion to and
respect for parents) and strong ties among family members
heavily influence end-of-life decisions.4,16 – 18 In Japan, many
identify themselves as both Buddhist and Shinto; they do not
attend worship services or participate in seasonal rituals for
several occasions without religious aims.4,19 – 22 In these 3
countries, the situation is rapidly changing, and clarification
of the common beliefs and practices in East Asian countries
when death is approaching is valuable for the provision of
culturally suitable end-of-life care.
www.md-journal.com | 1
Cheng et al
The primary aim of this study was to explore the potential
differences in common beliefs and practices when death is
approaching, based on physician reports from these East
Asian countries.
METHODS
This was a cross-sectional survey of palliative care special-
ists in Japan, Korea, and Taiwan. Measurement outcomes were
physician-perceived frequencies when patients were dying.
There was also an assumption that rough estimates of the
frequencies were not seriously different from direct observation
of patients.
Subjects and Procedures
We distributed a questionnaire to palliative care phys-
icians. Rationale for selecting palliative care physicians for
this study is that physicians are reasonably reliable source of
information without burden on patients and families. Further-
more, palliative care physicians as data sources was highly
feasible, as they experienced many patient deaths and this proxy
method is reliable and common in palliative care research.23,24
Due to differences in specialist registry and survey feasibility
among the countries, we decided to adopt the most feasible
methods based on the actual situation in each country. All the
palliative care physicians in the three countries work in
the hospitals.
In Japan, all 605 palliative care physicians certified by the
Japanese Society of Palliative Medicine before June 2012 were
recruited. Physicians’ names and affiliations were obtained
from the website of the Society, and questionnaires were
distributed by mail with 2 reminders (a total of three). No
reward was provided.
In Korea, due to the lack of a nationwide registry of
palliative care physicians, questionnaires were distributed via
3 methods. One was 5-spot surveys at academic congresses or
symposiums related to palliative medicine from October 2013
to January 2014, and a total of 97 responses were obtained. The
second was an e-mail survey, and a total of 32 responses were
obtained from a convenient sample of 110 palliative care
physicians through a local network. The third was an additional
hospital-based survey involving palliative care physicians
working at 3 hospitals, and 82 responses were obtained. A
small monetary reward was given for each response.
In Taiwan, all 578 palliative care physicians certified by
the Taiwan Academy of Hospice Palliative Medicine before
June 2012 were recruited. Physicians’ e-mail addresses were
obtained from the Academy, and questionnaires were distrib-
uted via e-mail with 2 reminders (a total of three). A small
reward was given to each participant completing the survey.
In all countries, responses to the questionnaire were
voluntary, and confidentiality was maintained throughout all
investigations and analyses. No identification numbers were
linked with the original data. The ethical and scientific validity
were approved by institutional review boards in each country.
Measurements
For measurement outcomes, we decided to explore 4 areas:
(1) reluctance to take part in end-of-life discussions, (2) role of
family members, (3) home death, and (4) circumstances sur-
rounding death. These measurement outcomes were developed
based on a systematic literature review regarding this topic,1–22
discussion among research groups, and preliminary in-depth
2 | www.md-journal.com
Medicine
Volume 94, Number 39, October 2015
interviews. Face validity was confirmed by a pilot testing on 10
palliative care clinicians using qualitative format, and after
several revisions all finally agreed questionnaires were under-
standable and appropriate in each country. The questionnaire
was simultaneously developed in Japanese, Korean, Taiwa-
nese, and English. For all questions, the physicians were asked
to rate the frequencies they experienced them on a Likert scale
of 1 (not at all), 1 (rarely), 2 (sometimes), 3 (often), to 4 (very
often).
Reluctance to take part in end-of-life discussions was
investigated using 3 items: (1) family members are reluctant
to discuss end-of-life issues until the patient is at the terminal
stage; (2) because ‘‘bad things happen after you say them out
loud,’’ the patient and family are reluctant to talk about a future
worsening in the patient’s condition; and (3) because ‘‘a bad life
is better than a good death,’’ the patient and family are reluctant
to talk about a future worsening in the patient’s condition.
The role of family members was investigated using 3
items: (1) when the physicians report ‘‘bad news,’’ they speak
to the oldest or head of the family first; (2) families recognize as
important that all family members are present at the last
moment; and (3) children want to do everything to fulfill ‘‘filial
piety.’’ These involve the core concepts of Confucianism, that is,
respect for older persons and the family bond.
Home death was investigated using 5 items: (1) because it
is believed that ‘‘the soul will not be able to return from the
hospital,’’ patients and their families want death to occur at
home; (2) because ‘‘death at home brings bad luck,’’ patients
and their families do not want death to occur at home; (3) for the
sake of ‘‘saving face,’’ patients and their families do not want
death to occur at home; (4) certain family members are expected
to care for the patient; and (5) patients ask someone who is not
their relative (eg, friend or neighbor) to take care of them instead
of their family.
Circumstances surrounding death were examined using 4
items: (1) patients want a religious person to visit them during
their terminal phase; (2) when the patient dies, family members,
including men, do not hesitate to express emotions (eg, crying
out); (3) family members clean the dead body of the patient
themselves; and (4) family members bathe the patient after
death. The last 2 were excluded in the Korean questionnaire
because a pilot test revealed that these questions made no sense
to Korean participants.
Statistical Analyses
To explore the frequencies of physician experience, we
adopted 2 comparison methods. One was the analysis of var-
iance (ANOVA) among the 3 countries; and another was
comparing the frequencies of often or very often using Chi-
square tests. Both analyses obtained the same results, so we
decided to provide the results of the latter, that is, frequencies of
the respondents who answered often or very often, for easier
interpretation. We calculated 95% confidence intervals, and a P-
value less than 0.01 was regarded as significant.
Specialty was categorized as internal medicine (general
internal medicine, subspecialties of internal medicine, psycho-
somatic medicine, and family practice), surgery (surgery and
related subspecialties, such as gynecology and otorhinolaryn-
gology), anesthesiology (anesthesiology, pain medicine), and
oncology (medical oncology, radiation oncology, and clinical
oncology).
All analyses were performed using the Statistical Package
for the Social Sciences (ver. 11.0) (Chicago, IL).
Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved.
Medicine
Volume 94, Number 39, October 2015 Cross-Cultural Study in East Asia

TABLE 1. Physicians’ Characteristics

Characteristics Korea (n ¼ 211) Japan (n ¼ 505) Taiwan (n ¼ 207) P


Sex <0.001
Male 105 (51%) 417 (83%) 138 (70%)
Female 103 (49%) 85 (17%) 59 (30%)

Specialty <0.001
Internal medicine 119 (57%) 147 (30%) 158 (80%)
Surgery 19 (9.0%) 183 (37%) 4 (2.0%)
Oncology 67 (32%) 52 (10%) 33 (17%)
Anesthesiology 4 (1.9%) 116 (23%) 3 (1.5%)

Clinical experience (mean years, standard deviation) 11.3 (8.3) 26.5 (7.5) 14.4 (8.9) <0.001
Working area <0.001y
Urban 136 (65%) 273 (55%) 137 (69%)
Rural 75 (36%) 228 (46%) 61 (31%)

Significant difference was observed in Japan versus Korea, Korea versus Taiwan, and Taiwan versus Japan (P < 0.001).
y
Japan versus Korea (P ¼ 0.016) and Taiwan versus Japan (P < 0.001).

RESULTS Role of Family Members

A total of 505 and 207 responses (83% and 36%, respect-
ively) were obtained from physicians recruited from Japan and
Taiwan, respectively. In Korea, a total of 211 responses were
obtained. Mean clinical experience of the physicians was more
than 10 years in all countries (Table 1).
Reluctance to Take Part in End-of-Life
Discussions
There was a similar trend in responses to reluctance to take
part in end-of-life discussions from the 3 countries: Japan was
the lowest, Korea was moderate, and Taiwan was the highest.
While 50% of the Japanese physicians reported that they often
or very often experienced reluctance on the part of family
members to discuss end-of-life issues, the corresponding figures
were 59% in Korea and 70% in Taiwan. Two specific reasons to
avoid end-of-life discussion, ‘‘bad things happen after you say
them out loud’’ and ‘‘a bad life is better than a good death’’
were most frequently observed in Taiwan, to a significant extent
(Figure 1).
Prioritizing the oldest or head of the family to break bad
news and all family members being present at the time of death
were significantly more frequently observed in Korea and
Taiwan than in Japan, without a statistically significant differ-
ence between Korea and Taiwan. More than 60% of Korean and
Taiwanese physicians reported they often or very often found
that bad news was given to the oldest or head of the family first,
while only 16% of Japanese physicians reported a similar
situation. More than 80% of Korean and Taiwanese physicians
reported they often or very often experienced that the presence
of all family members was regarded as important, while 50% of
Japanese physicians a reported similar situation. With regard to
the concept of filial piety, there was a significant trend of Japan
being the lowest, Korea moderate, and Taiwan having the
highest regard (Figure 2).
Home Death
As a reason for wanting home death, half of Taiwanese
physicians reported they often or very often experienced that
patients/family wanted the patient to die at home because the

 
FIGURE 1. Reluctance about end-of-life discussions. P < 0.01; FIGURE 2. Role of family members. P < 0.01; bars demonstrate
bars demonstrate 95% confidence intervals. 95% confidence intervals.

Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved. www.md-journal.com | 3
Cheng et al

FIGURE 3. Home death. P < 0.01; bars demonstrate 95% con-
fidence intervals.
soul would not be able to return from the hospital; while less
than 5% in the other countries had a similar experience
(Figure 3). There were no marked differences among the
countries in the remaining items: less than 15% of physicians
in all 3 countries stated they often or very often experienced
‘‘death at home brings bad luck’’ and ‘‘saving face’’ as reasons
to avoid home death. In all countries, more than 70% of the
physicians reported they often or very often found that certain
family members were expected to care for the patient at home,
but 10% or less found that friends and neighbors were expected
to help in this regard.
Circumstance Surrounding Death
At the time of death, a smaller number of Japanese
physicians than Korean or Taiwanese reported that they often
or very often experienced patients wanted a religious person to
visit and family members did not hesitate to express emotions
(crying out). On the other hand, about 40% of the Japanese
physicians reported that they often or very often experienced
family members cleaned the dead body of the patient by
themselves (Figure 4).

FIGURE 4. Circumstances surrounding death. P < 0.01; bars
demonstrate 95% confidence intervals.
4 | www.md-journal.com
Medicine
Volume 94, Number 39, October 2015
DISCUSSION
This is the first cross-cultural study investigating common
beliefs and practices when death is approaching in East Asian
countries. The clinical implication for a worldwide audience is
that this study clarified some important points every clinician
should pay attention to in caring for dying patients, from the
East Asian perspective. The research implication is that the
items investigated in this study can be used in further inter-
national surveys of other ethnic populations. Through more
studies on multicultural populations worldwide, we can gain
insight into what is usual in terms of beliefs and practices when
people die.
This study revealed that Confucianism has a strong influ-
ence on behaviors when people die in Korea and Taiwan, but not
in Japan. Confucian philosophy emphasizes family values and
respect for the oldest person. Filial piety is a virtue of respect for
parents, and is central to Confucian role ethics.4 In an end-of-
life context, filial piety means to take care of the parents and do
everything good for the parents. In this study, all items inves-
tigating family roles were rated higher in Korea and Taiwan,
that is, physicians give bad news to the oldest or head of the
family first, all family members are present at the last moment,
and sons/daughters want to do everything they can to esteem
filial piety. In Korea and Taiwan, dying and death is perceived
not as a personal issue, but rather as a family issue. This study
suggests that the Japanese situation seems to be a little different,
consistent with some observations.4,20 Japan is moving to put
more focus on the individual in the modern period, and the role
of the family is not as pronounced as in other East Asian
countries.
The role of religion is a focus in cross-cultural studies. This
study highlights the lack of participation of religious persons in
the dying process in Japan, in contrast to Korea and Taiwan.
Japanese generally identify themselves as Buddhist and Shinto,
but many do not actively participate in religious activities.4 In
contrast, 40% of Koreans identifies themselves as Buddhists or
Christians; Korea is an Asian country with a large Christian
population. Many Taiwanese identify themselves as Buddhist
and Taoist, and frequently visit temples to pray as a part of daily
life.4 This finding is in contrast to the situation in Japan, where
palliative care clinicians regard religion as less important for a
good death, and there are difficulties in obtaining a clear
consensus about what spiritual care is and how spiritual care
should be provided in Japan.22 Moreover, in this survey,
Japanese family members were reported to be less expressive
of emotions at the time of death, as they believe silence is a
preferable attitude at sad events. This finding is in contrast with
the Korean and Taiwanese attitude, in which honestly revealed
emotion is an expression of respect for the deceased.4,25 Clin-
icians should note that the extent to which family members
express emotion at the time of death depends on the cultural
norms of the society and varies culture to culture.
In addition to the effects of Confucianism, each country
has cultural issues unique to that country. In Taiwan, people
want to die at home because they believe the soul would not be
able to return from the hospital if the patient dies in the hospital.
This belief leads to a phenomenon of discharge just before
death: Taiwanese sometimes go back home just before death
even if the patient is unconscious, and this is legal in Taiwan.4
An international meta-analysis revealed factors associated with
death at home, but such beliefs were not investigated.26 This
finding suggests that, in some countries, death at home is not a
medical issue alone, but rather a cultural one. Also, cleaning the
dead body of the patient by family members themselves is a
Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved.
Medicine
Volume 94, Number 39, October 2015
Japanese tradition. Domestic nationwide surveys revealed that
some inpatient hospices in Japan include bathing the patient
after death as part of their usual care, but 40% of families
cleaned the dead body of the patient by themselves.27,28
With regard to traditional reasons to avoid home death, that
is, ‘‘death at home brings bad luck’’ and ‘‘saving face,’’ there
were no marked differences among the three countries. Also,
there were no differences in who were the expected caregivers
at home: in these 3 countries, those expected to be caregivers at
home were the family; friends, or neighbors were rarely
expected to be informal caregivers at home. Although saving
face or sekentei has been listed as main reason families in Japan
do not have professional caregivers,4 we found this value is
consistent in the East Asian countries investigated. Unlike some
countries, such as in Latin America, where friends or neighbors
are acknowledged as informal caregivers at home, the East
Asian situation puts a strong emphasis on family.1,2,4
Reluctance to participate in end-of-life discussions is a
common practice all over the world, especially in Asia.4 The
most hypothesized interpretation is that Confucianism does not
systemically refer to life after death, and death has been a taboo
for long periods.4 This study examined 2 specific reasons for the
unwillingness to take part in end-of-life discussions, that is,
‘‘bad things happen after you say them out loud’’ and ‘‘a bad
life is better than a good death.’’ These are traditional views,
especially frequently observed in Taiwan, and more frequently
in Taiwan than in Korea and Japan. Taiwan is the first county to
enact a Natural Death Act,20–22 but palliative care clinicians in
this study reported they experienced more reluctance among
patients and families in Taiwan to discuss end-of-life issues than
in Korea and Japan. This might lead to professional and ethical
conflicts or dilemmas among Taiwanese healthcare pro-
fessionals,20–22 and further follow-up study on changes in
public attitudes about end-of-life discussions in East Asia
is needed.
This study has several limitations. First, the response rate
in Taiwan was low, and, in Korea, there was no registry of the
subjects to be studied. Recruitment process including payment
was also different among the countries. Therefore, there may be
a considerable nonresponse and selection bias. Second, the
measurement instruments were not formally validated. Never-
theless, we believe that this is an acceptable limitation, because
there are no standardized measurement tools for our study aims
and the results obtained in the study are highly interpretable.
Third, this study investigated physicians’ views only, which
might be different from actual observations of patients and
families. Future studies should include direct observation of the
behaviors of patients and families when death is approaching.
Finally, this study did not explore the potential influences if
sociodemographic factors such as education and economic
status and outcomes were measured. This might be addressed
in future studies.
In conclusion, there are significant intercountry differ-
ences in beliefs and behaviors when death is approaching in
East Asia. Palliative care should be provided in a culturally
acceptable manner for each country. Future studies on direct
observations of patients and families are needed.
ACKNOWLEDGMENTS
We thank Ching-Yu Chen, MD (Department of Family
Medicine, College of Medicine & Hospital, National Taiwan
University), Chien-An Yao, MD (Department of Family Medi-
cine, National Taiwan University Hospital), Jaw-Shiun Tsai,
Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved.
Cross-Cultural Study in East Asia
MD (Department of Family Medicine, National Taiwan Uni-
versity Hospital) for the help in data collection and guidance in
this study. Taro Nishiura, Med (Division of Clinical Psychology,
Kyoto University), and Sayako Takahashi, Ph.D. (Division of
Clinical Psychology, Kyoto University) for the help in data
management, translation, and technical assistance in this study.
We thank the Japanese Hospice Foundation for the
financial support.
REFERENCES
1. Kagawa-Singer M, Blackhall LJ. Negotiating cross-cultural issues at
the end of life. JAMA. 2001;286:2993–3001.
2. Searight HR, Gafford J. Cultural diversity at the end of life: issues
and guidelines for family physicians. Am Fam Physician.
2005;71:515–522.
3. Gysels M, Evans N, Meñaca A, et al. Culture is a priority for
research in end-of-life care in Europe: A research agenda. J Pain
Symptom Manage. 2012;44:285–294.
4. Glass AP, Chen LK, Hwang E, et al. A cross-cultural comparison of
hospice development in Japan, South Korea, and Taiwan. J Cross
Cult Gerontol. 2010;25:1–19.
5. Blackhall LJ, Murphy ST, Frank G, et al. Ethnicity and attitudes
toward patient autonomy. JAMA. 1995;274:820–825.
6. Mystakidou K, Parpa E, Tsilila E, et al. Cancer information
disclosure in different cultural contexts. Support Care Cancer.
2004;12:147–154.
7. Bruera E, Neumann CM, Mazzocato C, et al. Attitudes and beliefs
of palliative care physicians regarding communication with termin-
ally ill cancer patients. Palliat Med. 2000;14:287–298.
8. Kwak J, Haley WE. Current research findings on end-of-life decision
making among racially or ethnically diverse groups. Gerontologist.
2005;45:634–641.
9. Raijmakers NJH, van Zuylen L, Costantini M, et al. Artificial
nutrition and hydration in the last week of life in cancer patients. A
systematic literature review of practices and effects. Ann Oncol.
2011;22:1478–1486.
10. Chen CH, Tang ST, Chen CH. Meta-analysis of cultural differences
in western and Asian patients-perceived barriers to managing cancer
pain. Palliat Med. 2011;26:206–221.
11. Yaguchi A, Truog RD, Curtis JR, et al. International differences in
end-of-life attitudes in the intensive care unit. Arch Intern Med.
2005;165:1970–1975.
12. Cuttini M, Nadai M, Kaminski M, et al. End-of-life decisions in
neonatal intensive care: physicians’ self-reported practices in seven
European countries. Lancet. 2000;355:2112–2118.
13. Chiu TY, Hu WY, Huang HL, et al. Prevailing ethical dilemmas in
terminal care for patients with cancer in Taiwan. J Clin Oncol.
2009;27:3964–3968.
14. Hu WY, Huang CH, Chiu TY, et al. Factors that influence the
participation of healthcare professionals in advance care planning for
patients with terminal cancer: a nationwide survey in Taiwan. Soc
Sci Med. 2010;70:1701–1704.
15. Cheng SY, Dy S, Hu WY, et al. Factors affecting the improvement
of quality of dying of terminally ill patients with cancer through
palliative care: a ten-year experience. J Palliat Med. 2012;15:854–
862.
16. Mo HN, Shin DW, Woo JH, et al. Is patient autonomy a critical
determinant of quality of life in Korea? End-of-life decision making
from the perspective of the patient. Palliate Med. 2011;26:222–231.
17. Yun YH, Lee CG, Kim SY, et al. The attitudes of cancer patients
and their families toward the disclosure of terminal illness. J Clin
Oncol. 2004;22:307–314.
www.md-journal.com | 5
Cheng et al
18. Yun YH, Lee MK, Kim SY, et al. Impact of awareness of terminal
illness and use of palliative care or intensive care unit on the
survival of terminal ill patients with cancer: prospective cohort
study. J Clin Oncol. 2011;29:2474–2480.
19. Ruhnke GW, Wilson SR, Akamatsu T, et al. Ethical decision making
and patient autonomy: a comparison of physicians and patients in
Japan and the United States. Chest. 2000;118:1172–1182.
20. Asai A, Fukuhara S, Lo B. Attitudes of Japanese and Japanese-American
physicians towards life-sustaining treatment. Lancet. 1995;346:356–359.
21. Matsumura S, Bito S, Liu H, et al. Acculturation of attitudes toward
end-of-life care: a cross-cultural survey of Japanese Americans and
Japanese. J Gen Intern Med. 2002;17:531–539.
22. Murata H, Morita T. Japanese Force Task. Conceptualization of
psycho-existential suffering by the Japanese Task Force: the first
step of a nationwide project. Palliat Support Care. 2006;4:279–285.
23. Teno JM. Measuring end-of-life care outcomes retrospectively. J
Palliat Med. 2005;8(Suppl 1):S42–S49.
6 | www.md-journal.com
Medicine
Volume 94, Number 39, October 2015
24. Cheng SY, Dy S, Huang SB, et al. Comparison of proxy ratings
of main family caregivers and physicians on the quality of dying
of terminally ill cancer patients. Jpn J Clin Oncol. 2013;43:
795–804.
25. Hendriks MC, Croon MA, Vingerhoets AJ. Social reactions to adult
crying: the help-soliciting function of tears. J Soc Psychol.
2008;148:22–41.
26. Gomes B, Higginson IJ. Factors influencing death at home in
terminally ill patients with cancer: systematic review. BMJ.
2006;332:515–521.
27. Shinjo T, Morita T, Miyashita M, et al. Care for the bodies of
deceased cancer inpatients in Japanese palliative care units. J Palliat
Med. 2010;13:27–31.
28. Yamawaki M, Morita T, Kiyohara E, et al. Experience and
evaluation of caring the body of deceased patient together with
nurses: a view from bereaved families. Palliat Care Res.
2015;10:101–107.
Copyright # 2015 Wolters Kluwer Health, Inc. All rights reserved.
Original Article

Impact of Aggressive Management and Palliative Care

on Cancer Costs in the Final Month of Life
Matthew C. Cheung, MD, SM1; Craig C. Earle, MD, MSc2,3; Jagadish Rangrej, MSc3; Thi H. Ho3; Ning Liu, MSc3;
Lisa Barbera, MD, MPA1,3; Refik Saskin, MSc3; Joan Porter, MSc3; Soo Jin Seung, Hon BSc4; and Nicole Mittmann, PhD4,5,6,7

BACKGROUND: A significant share of the cost of cancer care is concentrated in the end-of-life period. Although quality measures of
aggressive treatment may guide optimal care during this timeframe, little is known about whether these metrics affect costs of care.
METHODS: This study used population data to identify a cohort of patients who died of cancer in Ontario, Canada (2005-2009). Indi-
viduals were categorized as having received or having not received aggressive end-of-life care according to quality measures related
to acute institutional care or chemotherapy administration in the end-of-life period. Costs (2009 Canadian dollars) were collected
over the last month of life through the linkage of health system administrative databases. Multivariate quantile regression was used
to identify predictors of increased costs. RESULTS: Among 107,253 patients, the mean per-patient cost over the final month was
$18,131 for patients receiving aggressive care and $12,678 for patients receiving nonaggressive care (P <.0001). Patients who received
chemotherapy in the last 2 weeks of life also sustained higher costs than those who did not (P <.0001). For individuals receiving end-
of-life care in the highest cost quintile, early and repeated palliative care consultation was associated with reduced mean per-patient
costs. In a multivariate analysis, chemotherapy in the 2 weeks of life remained predictive of increased costs (median increase, $536;
P <.0001), whereas access to palliation remained predictive for lower costs (median decrease, $418; P <.0001). CONCLUSIONS: Can-
cer patients who receive aggressive end-of-life care incur 43% higher costs than those managed nonaggressively. Palliative consulta-
tion may partially offset these costs and offer resultant savings. Cancer 2015;121:3307-15. V C 2015 The Authors. Cancer published by

Wiley Periodicals, Inc. on behalf of American Cancer Society. This is an open access article under the terms of the Creative Commons
Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is prop-
erly cited, the use is non-commercial and no modifications or adaptations are made.

KEYWORDS: chemotherapy, costs, end-of-life care, palliative care, quality measures.

INTRODUCTION
The economic burden of cancer is high for both individuals and society as a whole. In 2009, the National Institutes of
Health estimated the annual costs of cancer to have reached 216.6 billion dollars, with 86.6 billion dollars representing
direct medical costs of cancer care.1 Because of the changing incidence, prevalence, and outcomes of malignant diseases,
the financial consequences are anticipated to increase further.2 In particular, a disproportionate share of health care costs is
concentrated in the end-of-life period.2-4 Prior research in the United States and Canada has established quality measures
of aggressive management for the last months of life.5-9 High rates of unplanned medical encounters such as emergency
room visits, hospitalizations, and intensive care stays, for example, may indicate inattention to symptomatic issues, a lack
of advance directives, or inadequate utilization of home and hospice services. Chemotherapy administered in the last days
of life may offer little chance of clinical benefit but notable consequences, including toxicities that impair quality of life
and delay access to hospice care.10,11 Although quality measures of aggressive management may signal suboptimal oncol-
ogy care, little is known about whether quality-based end-of-life care can result in cost savings.
Implicit in the assumption that aggressive care is suboptimal at the end of life is the suggestion that a palliative focus
could improve the quality of care. When physicians simply engage in end-of-life conversations, their patients experience
less aggressive medical care near death (including fewer intensive care unit [ICU] admissions) with improved quality of
life.12,13 With additional consultation with palliative care teams, further benefits in patient satisfaction and health care

Corresponding author: Matthew C. Cheung, MD, SM, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, T2-031, Toronto,
Ontario, Canada M4N3M5; Fax: (416) 480-6002; matthew.cheung@sunnybrook.ca
1
Odette Cancer Centre, Sunnybrook Health Sciences Centre and University of Toronto, Toronto, Canada; 2Ontario Institute for Cancer Research, Toronto, Canada;
3
Institute for Clinical Evaluative Sciences, Toronto, Canada; 4Health Outcomes and Pharmacoeconomic Research Centre, Sunnybrook Research Institute,
Sunnybrook Health Sciences Centre, Toronto, Canada; 5Canadian Centre for Applied Research in Cancer Control, Toronto, Canada; 6Department of Pharmacology,
University of Toronto, Toronto, Canada; 7International Centre for Health Innovation, Richard Ivey Business School, Western University, London, Canada.

Additional Supporting Information may be found in the online version of this article.

DOI: 10.1002/cncr.29485, Received: March 17, 2015; Revised: April 22, 2015; Accepted: April 24, 2015, Published online May 29, 2015 in Wiley Online Library
(wileyonlinelibrary.com)

Cancer September 15, 2015 3307

Original Article
TABLE 1. Administrative Databases
Database
Ontario Cancer Registry
Ontario Health Insurance Plan
Canadian Institute for Health Information Discharge Abstract Database
National Ambulatory Care Reporting System
Home Care Database
Registered Persons Database
utilization have been demonstrated.14,15 As such, it would
be anticipated that palliation could reduce the use of
costly resources. Supporting this contention, preliminary
research has suggested that end-of-life discussions16 and
conversations about spirituality17 might reduce costs of
care. Conversely, palliative consultation and placements
have their own intrinsic costs. The relation between pallia-
tive care, aggressiveness of care, and costs warrants further
evaluation.
We used population-level administrative databases
to identify the costs of end-of-life care in patients man-
aged with aggressive and nonaggressive intent on the basis
of previously established quality measures. Distinctly from
the aggressiveness of care received, we further assessed the
impact of preceding palliative care services on costs of
care. We hypothesized that patients receiving aggressive
care near the end-of-life would use more resources and
correspondingly incur greater costs of care in comparison
with individuals who did not receive aggressive care.
MATERIALS AND METHODS
Study Design
We used public administrative databases to identify
resources and costs associated with end-of-life cancer care
for adult patients in Ontario, Canada who died between
January 1, 2005 and December 31, 2009. We excluded
cases if they did not have a valid provincial health insur-
ance number, died within 30 days of the initial cancer di-
agnosis, or were younger than 20 years. Individuals in the
decedent cohort were categorized as having received or
having not received aggressive end-of-life management
according to the following quality measures5,7,18: 1)
chemotherapy administered within 14 days of death, 2)
more than 1 emergency department (ED) visit within 30
days of death, 3) more than 1 hospitalization within 30
days of death, or 4) at least 1 ICU admission within 30
3308
Description
Population-based database with approximately 95% of all provincial cancer
diagnoses captured
Contains all billing claims for physician services and certain other health
professionals (including fee codes for chemotherapy administration)
Contains demographic, administrative, and clinical data related to hospital
admissions (including the provision of intensive care unit care)
Contains information on ambulatory care visits (including emergency depart-
ment care)
Contains information on government-funded care at home
Contains basic demographics and dates of death for all Ontario residents el-
igible for Ontario Health Insurance coverage
days of death. This study was approved by the Research
Ethics Board at Sunnybrook Health Sciences Centre.
Cohort Selection and Variable Definitions
Patients with cancer as the cause of death were identified
in the Ontario Cancer Registry.19 International Classifica-
tion of Diseases, Ninth Revision (ICD9) codes were used to
identify the type of cancer death (breast [ICD9-174],
lung [ICD9-162], colorectal [ICD9-153 and ICD9-154],
prostate [ICD9-185], hematologic [ICD9-196 and
ICD9-200 to ICD9-208], and other [all other ICD9 can-
cer codes]). As previously described, patient cases from
the Ontario Cancer Registry were linked to administrative
databases (Table 1) with encrypted insurance numbers to
determine whether quality measures were performed.5
The neighborhood income quintile was assigned by
the linkage of the patient’s postal code (from the Regis-
tered Persons Database) to the median income by dissemi-
nation area identified from the Statistics Canada 2006
census. The rural status (communities with a population
size < 10,000) and the regional designation in Ontario
(based on the local health integration networks) were
based on the city, postal code, and dissemination area of
the patient’s residence. Patient comorbidity was calculated
with the Deyo adaptation20 of the Charlson comorbidity
index measure21 according to inpatient diagnoses in the
last 2 years of life.
Palliative care was defined by hospitalization, home
care, or physician billing codes specific for palliative con-
sultation (see online supporting information). To identify
patients who had received repeated and earlier palliative
care services before the final days of life, we defined
patients as palliative if at least 2 palliative care codes were
identified at least 30 days apart and in the last 365 days
before death. The palliative care designation was a unique
predictor variable mutually distinct from the aggressive
quality measures.
Cancer September 15, 2015
 Cost of Aggressive End-of-Life Care/Cheung et al

TABLE 2. Baseline Demographic and Clinical Characteristics

Patients Not Experiencing Any Patients Experiencing Any Total

Characteristic Aggressive Care (n 5 83,158) Aggressive Care (n 5 24,095) (n 5 107,253)

Age, mean 6 SD, y 71.92 6 12.84 68.14 6 12.85 71.07 6 12.94

Female, No. (%) 40,832 (49.1) 10,329 (42.9) 51,161 (47.7)
Charlson score per point, mean 6 SD 5.11 6 2.65 5.63 6 2.41 5.23 6 2.61
Duration of disease, mean 6 SD, y 3.43 6 5.28 3.17 6 5.05 3.37 6 5.23
Cancer type, No. (%)
Breast 6976 (8.4) 1950 (8.1) 8926 (8.3)
Colorectal 9715 (11.7) 2464 (10.2) 12,179 (11.4)
Hematologic 6763 (8.1) 3399 (14.1) 10,162 (9.5)
Lung 19,678 (23.7) 5937 (24.6) 25,615 (23.9)
Prostate 5196 (6.2) 1147 (4.8) 6343 (5.9)
Other 34,830 (41.9) 9198 (38.2) 44,028 (41.1)
Income quintile, No. (%)
Highest 15,634 (18.8) 4392 (18.2) 20,026 (18.7)
Second highest 15,627 (18.8) 4530 (18.8) 20,157 (18.8)
Middle 15,945 (19.2) 4782 (19.8) 20,727 (19.3)
Second lowest 17,719 (21.3) 5291 (22.0) 23,010 (21.5)
Lowest 17,763 (21.4) 5004 (20.8) 22,767 (21.2)
Rural residence, No. (%) 11,503 (13.8) 4484 (18.6) 15,987 (14.9)
Region of residence, No. (% by LHIN)
01. Erie–St. Clair 4879 (5.9) 1437 (6.0) 6316 (5.9)
02. South West 6835 (8.2) 2169 (9.0) 9004 (8.4)
03. Waterloo 4328 (5.2) 1150 (4.8) 5478 (5.1)
04. Hamilton 11,530 (13.9) 2893 (12.0) 14,423 (13.4)
05. Central West 3024 (3.6) 1075 (4.5) 4099 (3.8)
06. Mississauga 5404 (6.5) 1439 (6.0) 6843 (6.4)
07. Toronto 7228 (8.7) 1707 (7.1) 8935 (8.3)
08. Central 8075 (9.7) 2532 (10.5) 10,607 (9.9)
09. Central East 9290 (11.2) 2973 (12.3) 12,263 (11.4)
10. South East 4174 (5.0) 1268 (5.3) 5442 (5.1)
11. Champlain 8298 (10.0) 2017 (8.4) 10,315 (9.6)
12. North Simcoe 3181 (3.8) 1210 (5.0) 4391 (4.1)
13. North East 4937 (5.9) 1682 (7.0) 6619 (6.2)
14. North West 1804 (2.2) 533 (2.2) 2337 (2.2)
Designated palliative, No. (%) 56,674 (68.2) 13,041 (54.1) 69,715 (65.0)
Quality measures, No. (%)
>1 ED visit in last 30 d of life 0 (0.0) 15,812 (65.6) 15,812 (14.7)
Chemotherapy in last 14 d of life 0 (0.0) 3893 (16.2) 3893 (3.6)
ICU admission in last 30 d of life 0 (0.0) 5923 (24.6) 5923 (5.5)
>1 hospitalization in last 30 d of life 0 (0.0) 6765 (28.1) 6765 (6.3)

Abbreviations: ED, emergency department; ICU, intensive care unit; LHIN, local health integration network; SD, standard deviation.

Cost Calculation aggressive intervention. Crude rates for each indicator of

We used database linkages to identify and determine the
costs of all health-related resources used by patients
over the last month of life and paid for by the Ontario
Ministry of Health and Long-Term Care (see online sup-
porting information). Costs were calculated for acute hos-
pitalizations, ED visits, ambulatory oncology visits, same-
day surgery, complex continuing care (inclusive of pallia-
tive care placement), long-term care, home care, physician
assessments, certain intravenous chemotherapies, oral
medications, and other resources (mental health admis-
sions, dialysis, rehabilitation, and devices).
Statistical Analysis
We examined baseline descriptive characteristics for
patients who experienced at least 1 indicator of aggressive
management versus those who did not experience any
aggressive end-of-life care were computed per year from
2005 to 2009, and they were stratified according to the
palliative designation.
Total and disaggregated mean costs per patient were
presented descriptively for patients who experienced at
least 1 indicator of aggressive care and for those who did
not. Because of the higher intrinsic costs associated with
the hospitalization, ED, and ICU quality measures, we
separately analyzed the costs for those who received chem-
otherapy in the last 14 days of life and those who did not.
Analyses were further stratified according to whether
patients had a palliative designation.
Because of the skewness in the cost data and expo-
nential characteristics of the distribution, univariate quan-
tile regression was used to test the significance of the cost

Cancer September 15, 2015 3309

Original Article

Figure 1. Trends in aggressive end-of-life care in Ontario according to the palliative designation (% of cohort on y-axis): (A) >1
ED visit in the last 30 days of life, (B) chemotherapy in the last 14 days of life, (C) intensive care unit admissions in the last 30
days of life, and (D) hospitalizations in the last 30 days of life. ED indicates emergency department.

differential between aggressive and nonaggressive end-of-
life care. Analyses were further stratified according to the
palliative designation. Multivariate quantile regression
analysis was used to estimate effects of the palliative desig-
nation and the aggressive quality measures on median
costs. In quantile regression, the quantiles or percentiles
of the distribution of costs are modeled as a function of
predictor characteristics.22 Because mean costs are poten-
tially more susceptible to outliers than median costs,23 we
felt that this modeling was a more appropriate method to
avoid overestimating the impact of the aggressive indicators
on overall end-of-life costs. Models were adjusted for the
age at death (continuous variable centered on the age of 20
years), sex, cancer type (hematologic vs nonhematologic),
income quintile, Charlson score, duration of disease, resi-
dence (urban vs rural), and year of death (continuous vari-
able centered on the year 2005). All analyses were
performed with SAS 9.1.3 (SAS Institute, Cary, NC).
RESULTS
Baseline Characteristics and Aggressive
Care Indicators
There were 107,253 patients who died of cancer in On-
tario, Canada and met our inclusion criteria to form the
cohort for the study. Table 2 presents demographic char-
acteristics of patients who did and did not receive aggres-
sive care. In the cohort, 24,095 patients (22.5%)
experienced at least 1 quality measure of aggressive end-
of-life care.
During the study period, 3.6% of the cohort
received systemic chemotherapy in the last 2 weeks of life.
The remaining indicators were noted more frequently:
14.7% experienced more than 1 ED visit in the last 30
days of life, 6.3% of patients experienced multiple hospi-
talizations in the final month of life, and 5.5% experi-
enced ICU admissions in the final month of life. Figure 1
depicts the trends in aggressive management for each indi-
cator over the study time horizon stratified by the pallia-
tive designation. At all time points, patients who were
designated as palliative were less likely to receive aggressive
care according to all 4 quality measures.
Costs of Care
The mean cost per patient in the last month of life was
$13,903 Canadian dollars (median, $10,859; interquar-
tile range [IQR], $5387-$18,498). The predominant cost
driver during this period was hospitalizations (contribut-
ing 64.7% of the total cost), with a mean cost of $8998.

3310 Cancer September 15, 2015


TABLE 3. Mean Costs for Aggressive and
Nonaggressive Care (in Canadian dollars)
Nonaggressive
Care
Hospitalizations 7971 (12,483)
Emergency department visits 214 (325)
Outpatient (same-day) surgery 17 (151)
Outpatient cancer clinic 237 (787)
Complex continuing care 1046 (2951)
Long-term care 206 (718)
Home care costs 1370 (2229)
Chemotherapy (NDFP) 28 (253)
Oral/outpatient drugs 401 (1089)
Physician billings 1060 (1501)
Other costsa 127 (1912)
Total costs 12,678 (12,754)
Abbreviation: NDFP, New Drug Funding Program.
Disaggregated costs may not add up to the total amount because of
rounding. Data are presented as means and standard deviations.
a
Other costs include admissions for mental health, dialysis, rehabilitation,
and devices.
Table 3 lists the overall mean costs and disaggre-
gated costs for patients who received aggressive care and
patients who did not. The mean per-patient costs over the
final month were $18,131 (median, $14,464; IQR,
$9437-$21,658) for patients receiving aggressive care and
$12,678 (median, $9586; IQR, $4456-$17,414) for
patients receiving nonaggressive care; the difference was
$5453 (P < .0001). For individuals with 1 (n 5 16,867),
2 (n 5 6220), 3 (n 5 946), or all 4 indicators of care
(n 5 62), the mean per-patient costs increased to
$17,001, $20,304, $23,690, and $22,700, respectively.
Aggressive care patients incurred higher costs for hos-
pitalization, ED visits, and physician services in
comparison with nonaggressive care patients. In contrast,
nonaggressive care was associated with higher costs for
the provision of complex continuing care (including inpa-
tient palliation) and home care. Hospitalizations remained
the key cost drivers for patients managed both aggressively
and nonaggressively ($12,541 and $7971, respectively).
We separately compared costs for individuals who
received systemic chemotherapy in the last 14 days of life
as the sole indicator of aggressive care. The mean per-
patient costs over the final month were $16,110 (median,
$12,433; IQR, $8045-$19,599) for patients receiving
chemotherapy and $13,820 (median, $10,794; IQR,
$5269-$18,454) for patients who did not receive treatment
(P < .0001). Patients treated with chemotherapy in the 2
weeks before death incurred higher hospitalization, ED,
physician-related, and chemotherapy drug costs, whereas
those who did not receive this treatment incurred higher
palliative care and home care costs (data not shown).
Cancer September 15, 2015
Cost of Aggressive End-of-Life Care/Cheung et al
Aggressive
Care
12,541 (14,032)
826 (530)
43 (251)
586 (1318)
273 (1259)
75 (436)
988 (1306)
170 (699)
454 (1044)
2041 (2201)
133 (1841)
18,131 (15,065)
Figure 2. Effect of the palliative designation according to
quintiles of end-of-life costs (in Canadian dollars) stratified
by patients receiving aggressive care and patients receiving
nonaggressive care.
In the overall cohort, 69,715 patients (65.0%) were
designated as palliative on the basis of system billing codes
associated with the provision of palliative services. For
patients ultimately managed aggressively and patients
managed nonaggressively, the proportions of patients ini-
tially designated as palliative were 54.1% and 68.2%,
respectively. Palliative patients incurred lower mean costs
(mean, $13,503 vs $14,644) but higher median costs
(median, $10,988 vs $10,573) in comparison with
patients who were not palliative. We further analyzed the
relation of the palliative designation with quintiles of
costs. For individuals receiving end-of-life care in the
highest cost quintiles (the top 2 quintiles), palliative care
consultation was associated with reduced mean per-
patient costs. However, for individuals receiving end-of-
life care in the lowest cost quintiles (the bottom 2 quin-
tiles), palliation was associated with increased per-patient
costs. This relation between palliative care and costs was
seen in patients who were managed either aggressively or
nonaggressively. The most significant reduction in mean
per-patient costs associated with the palliative designation
occurred in aggressively managed patients in the highest
cost quintile (Fig. 2).
Multivariate quantile regression models were con-
structed to study the effects of the palliative designation,
aggressive care indicators, and patient variables on median
end-of-life costs (Table 4). In the initial model including
the palliative designation, individuals were more likely to
have high end-of-life expenditures if they were younger,
were female, were in the lowest income quintile, had been
diagnosed with a hematologic malignancy, were affected
3311
3312
TABLE 4. Quantile Multivariate Regression to Estimate the Effects of Predictor Variables on Median End-of-Life Costs (in Canadian dollars)
Original Article

Model 1 Model 2 Model 3 Model 4 Model 5

Parameter Level Estimate (SE) P Estimate (SE) P Estimate (SE) P Estimate (SE) P Estimate (SE) P

Intercept 6562.99 (187.96) <.0001 5991.87 (185.15) <.0001 6509.6 (182.62) <.0001 5902.1 (186.42) <.0001 5651.37 (178.96) <.0001
Agea 264.22 (2.5) <.0001 258.71 (2.44) <.0001 263.41 (2.44) <.0001 255.87 (2.51) <.0001 252.93 (2.4) <.0001
b
Sex Male — .0447 — .0005 — .0287 — .0068 — .0006
Female 122.3 (60.93) 192.53 (55.19) 127.66 (58.36) 165.58 (61.22) 193.8 (56.57)
Year of deathc 540.35 (21.86) <.0001 523.55 (20.29) <.0001 540.42 (20.81) <.0001 515.36 (21.97) <.0001 544.43 (20.33) <.0001
Income quintile Highestb — .1696 — .2253 — .1703 — — .342
Second highest 135.62 (98.73) .0389 108.37 (89.37) .0587 131.39 (95.82) .0484 146.82 (99.31) .1393 87.44 (92.02) .1345
Middle 204.26 (98.88) .0864 153.43 (81.17) .4671 188.09 (95.28) .0863 201.37 (94.83) .0337 130.92 (87.48) .3506
Second lowest 159.61 (93.09) .0002 61.44 (84.48) .0006 149.36 (87.07) .0002 168.57 (91.51) .0655 79.54 (85.22) .0004
Lowest 368.81 (98.97) 297.01 (86.49) 365.75 (96.94) 404.14 (97.7) <.0001 311.04 (87.31)
Rurality Ruralb — .5717 — .0093 — .6169 — .5622 — .0027
Urban 47.67 (84.28) 214.26 (82.34) 41.65 (83.26) 48.47 (83.63) 236.68 (78.96)
Cancer type Nonhematologicb — <.0001 — <.0001 — <.0001 — <.0001 — <.0001
Hematologic 3749.67 (152.93) 3554.97 (161.81) 3728.2 (151.04) 3054.94 (156.95) 3333.51 (159.66)
Duration of cancer — <.0001 — <.0001 — <.0001 — <.0001 — <.0001
223.94 (4.83) 225.74 (2.7) 223.59 (4.05) 226.31 (4.97) 224.82 (3.6)
Charlson score 0b — — — — —
1 1543.29 (192.49) <.0001 1436.88 (166.97) <.0001 1550.24 (193.89) <.0001 1443.00 (169.11) <.0001 1525.72 (175.7) <.0001
2 7045.36 (58.42) <.0001 6609.97 (54.34) <.0001 7039.43 (55.24) <.0001 6697.68 (60.63) <.0001 6517.95 (56.61) <.0001
Palliative designation Nob — <.0001 — <.0001 — <.0001 — .6155 —
Yes 2423.29 (65.45) 2236.50 (60.14) 2418.26 (62.21) 233.21 (66.11) 2221.14 (61.37) .0003
>1 ED visit Nob — <.0001
Yes 2636.41 (73.15)
Chemotherapy indicator Nob — <.0001
Yes 536.35 (89.26)
ICU indicator Nob — <.0001
Yes 9546.38 (258.61)
>1 hospitalization Nob — <.0001
Yes 6077.94 (102.11)

Abbreviations: ED, emergency department; ICU, intensive care unit; SE, standard error.

Cancer
a
Continuous variable centered on the age of 20 years.
b
Reference group.
c
Continuous variable centered on the year 2005.

September 15, 2015


by multiple comorbidities, had experienced a shorter du-
ration of disease, or had died in the later years of the study
period (for a model example, see online supporting infor-
mation). The palliative designation independently
reduced median costs by $423 (P < .0001). In subsequent
models, the impact of the individual quality measures was
studied (models 2-5). In the model that incorporated
chemotherapy in the last 14 days of life, this aggressive in-
dicator was independently associated with increased costs
(median increase, $536; P < .0001), whereas the palliative
designation remained predictive of decreased costs (me-
dian decrease, $418; P < .0001).
DISCUSSION
Quality measures for the end-of-life period were initially
derived to determine health care systems that might strug-
gle in providing appropriate patient-centered care.7
Although aggressive care may be warranted or requested
for individual patients, on a system level, the associated
indicators of such care might denote inadequate prepara-
tion or a lack of supportive services for those in the last
days of life.24 These indicators have thus been positioned
as system measures for poor-quality care. Our results indi-
cate that aggressive end-of-life care is more expensive than
nonaggressive care. In the final month of life, the care for
aggressively managed patients was $5453 more (or 43%
costlier) than nonaggressive care. Aggressively managed
patients were more likely to incur higher costs for acute
care and physician services, whereas nonaggressively man-
aged patients generated higher costs for home care and
complex continuing care, which included the provision of
inpatient palliation.
Because 3 of the 4 quality measures were intrinsi-
cally linked with more expensive resource utilization,
namely, institutional care, we separately evaluated the
costs with the chemotherapy indicator as the sole
reflection of aggressive end-of-life care. Once again,
patients treated aggressively according to this definition
incurred higher costs than their nonaggressively treated
counterparts.
Our study is the first to examine the unique relation
of a palliative designation with aggressive management
and end-of-life expenditures. We chose a palliative desig-
nation to reflect repeated and earlier access to palliation.
For patients destined to receive more costly (and aggres-
sive) care, we found that earlier access to palliative consul-
tation and services could meaningfully reduce costs. For
patients who would ultimately consume few health care
resources (and potentially less aggressive care), palliative
care was associated with higher costs, likely because of the
Cancer September 15, 2015
Cost of Aggressive End-of-Life Care/Cheung et al
intrinsic costs required for such care. The palliative desig-
nation remained predictive for decreased costs in multi-
variate models after adjustments for clinical variables.
Notably, the sole model in which the cost savings with
palliation did not achieve statistical significance (median
decrease, $33; P 5 .6155) was the model in which the
ICU indicator dramatically increased costs (median
increase, $9546; P < .0001). In the remaining analyses,
the palliative designation continued to independently pre-
dict lower costs. This is in keeping with prior research
demonstrating less intensive care and resource consump-
tion when earlier palliative care is instituted.14 In random-
ized trials of patients with advanced cancer, patients
allocated to earlier palliative care reported improved qual-
ity of life and were less likely to receive aggressive end-of-
life care.15,25,26 In an observational study, Zhang et al16
studied the impact of end-of-life conversations between
cancer patients and physicians on costs in the last week of
life. Patients who reported discussions about palliation
incurred costs of $1876 (2008 US dollars), whereas the
cost was $2917 for those who did not. We speculate that
our palliative designation might serve as a surrogate for
conversations about end-of-life preferences or the avail-
ability of palliative services and might potentially modu-
late medical expenditures even among those destined to
receive aggressive care. Further research is needed to
define the optimal timing and nature of palliative consul-
tation to optimize resource utilization and, more impor-
tantly, patient outcomes.
Our results are in keeping with prior costing efforts
for cancer patients at the end of life in both public27 and
private health care systems.4 In the United States, Chastek
et al4 used claims data to identify cancer-related costs for
28,530 patients with active cancer 6 months before death.
They established costs related to inpatient stays, hospice
care, and outpatient services of $74,212 (2009 US dollars)
and identified inpatient care as the key cost driver. This
analysis focused on a select population of commercially
insured individuals and did not stratify outcomes accord-
ing to aggressive indicators of care. In contrast, a study
from Kyoto, Japan did analyze costs according to select
aggressive quality measures, albeit in a smaller population
(n 5 3143) of predominantly self-employed, unem-
ployed, or elderly individuals (with the majority older
than 75 years).28 The investigators studied the costs asso-
ciated with the last 3 months of life but exclusively for
those who received institutional care. Death in an acute-
care hospital (rate ratio [RR], 1.32; P < .001) and chemo-
therapy in the last month of life (RR, 1.25; P < .001)
were associated with higher end-of-life costs. In contrast
3313
Original Article
to our results, palliative consultation did not abrogate
costs (RR, 0.98; P 5 .66). This inconsistency may be at-
tributable to differences in our definition of palliation;
earlier and repeated access to palliative care consultation
might be required to have a downstream impact on
resource use and costs. Our research further builds on
these data by calculating a broad range of cancer and non–
cancer-related costs in a large, unrestricted, and diverse
cohort that includes the entire population of Ontario.
It is important to note potential limitations of this
study. The quality measures of aggressive management are
unable to consider patient preferences or define optimal
care on an individual basis. There may be occasions when
aggressive care is most appropriate and patient-centered.
Our results are subject to the limitations of the individual
data sets used and the biases inherent to a retrospective
design. In particular, we appreciate that not all relevant
societal costs (eg, lost productivity and out-of-pocket
costs) were captured, and they can impart a substantial fi-
nancial burden borne by patients and their caregivers.29,30
Finally, our results are based on the public insurance sys-
tem, and their generalizability to private or mixed systems,
as found in the United States, should be considered. It is
likely that the cost implications related to aggressive end-
of-life care are considerable in the United States. A parallel
increase in aggressive end-of-life care over time has been
documented in both countries. Moreover, the absolute in-
tensity of care, reflected by higher utilization of chemo-
therapy and ICU care near the end of life, is notably
higher in the United States versus Canada.5 Differences in
physician and hospital remuneration may further drive
this variability; for example, financial incentives for physi-
cians to prescribe chemotherapy do not exist in Canada,
and the public-payer system may limit certain aspects of
aggressive care, including the provision of chemotherapy
for refractory disease.24,31,32 Ultimately, the goal of
improving the quality of end-of-life care, increasing the
availability of palliative care, and optimizing the distribu-
tion of limited resources is a universal pursuit common to
all modern health care systems.
This study established the costs associated with
the end-of-life period according to aggressive measures
of care. Costs are substantially higher for patients man-
aged aggressively in the final weeks of life, and they are
driven by a heavy dependence on acute institutional
care. Health care systems that seek to optimize the
costs and quality of end-of-life care should explore the
possibility that earlier palliation might temper the
tendency toward aggressive management and ultimately
offset costs.
3314
FUNDING SUPPORT
This work was funded in part by the National Cancer Institute
(grant R01 CA91753-02) and an Ontario Institute for Cancer
Research Health Services Program Grant.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosure.
REFERENCES
1. National Heart, Lung, and Blood Institute. Fact Book Fiscal Year
2012. http://www.nhlbi.nih.gov/files/docs/factbook/FactBook2012.
pdf. Accessed March 17, 2015.
2. Tangka FK, Trogdon JG, Richardson LC, et al. Cancer treatment
cost in the United States: has the burden shifted over time? Cancer.
2010;116:3477-3484.
3. Fireman BH, Quesenberry CP, Somkin CP, et al. Cost of care for
cancer in a health maintenance organization. Health Care Financ
Rev. 1997;18:51-76.
4. Chastek B, Harley C, Kallich J, et al. Health care costs for patients
with cancer at the end of life. J Oncol Pract. 2012;8:75s-80s.
5. Ho TH, Barbera L, Saskin R, et al. Trends in the aggressiveness of
end-of-life cancer care in the universal health care system of Ontario,
Canada. J Clin Oncol. 2011;29:1587-1591.
6. Barbera L, Paszat L, Chartier C. Indicators of poor quality end-of-
life cancer care in Ontario. J Palliat Care. 2006;22:12-17.
7. Earle CC, Park ER, Lai B, et al. Identifying potential indicators of
the quality of end-of-life cancer care from administrative data. J Clin
Oncol. 2003;21:1133-1138.
8. Barnato AE, McClellan MB, Kagay CR, et al. Trends in inpatient
treatment intensity among Medicare beneficiaries at the end of life.
Health Serv Res. 2004;39:363-375.
9. Nappa U, Lindqvist O, Rasmussen BH, et al. Palliative
chemotherapy during the last month of life. Ann Oncol. 2011;22:
2375-2380.
10. Harrington SE, Smith TJ. The role of chemotherapy at the end of
life: "when is enough, enough?". JAMA. 2008;299:2667-2678.
11. Saito AM, Landrum MB, Neville BA, et al. The effect on survival of
continuing chemotherapy to near death. BMC Palliat Care. 2011;10:14.
12. Wright AA, Zhang B, Ray A, et al. Associations between end-of-life
discussions, patient mental health, medical care near death, and care-
giver bereavement adjustment. JAMA. 2008;300:1665-1673.
13. Mack JW, Cronin A, Keating NL, et al. Associations between end-
of-life discussion characteristics and care received near death: a pro-
spective cohort study. J Clin Oncol. 2012;30:4387-4395.
14. Howie L, Peppercorn J. Early palliative care in cancer treatment: ra-
tionale, evidence and clinical implications. Ther Adv Med Oncol.
2013;5:318-323.
15. Greer JA, Pirl WF, Jackson VA, et al. Effect of early palliative care
on chemotherapy use and end-of-life care in patients with metastatic
non–small-cell lung cancer. J Clin Oncol. 2012;30:394-400.
16. Zhang B, Wright AA, Huskamp HA, et al. Health care costs in the
last week of life: associations with end-of-life conversations. Arch
Intern Med. 2009;169:480-488.
17. Balboni T, Balboni M, Paulk ME, et al. Support of cancer patients’
spiritual needs and associations with medical care costs at the end of
life. Cancer. 2011;117:5383-5391.
18. Grunfeld E, Lethbridge L, Dewar R, et al. Towards using administra-
tive databases to measure population-based indicators of quality of end-
of-life care: testing the methodology. Palliat Med. 2006;20:769-777.
19. Clarke EA, Marrett LD, Kreiger N. Cancer registration in Ontario:
a computer approach. IARC Sci Publ. 1991;95:246-257.
20. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity
index for use with ICD-9-CM administrative databases. J Clin Epi-
demiol. 1992;45:613-619.
21. Charlson ME, Pompei P, Ales KL, et al. A new method of classify-
ing prognostic comorbidity in longitudinal studies: development and
validation. J Chronic Dis. 1987;40:373-383.
Cancer September 15, 2015

22. Koenker R, Bassett G. Regression quantiles. Econometrica. 1978;46:
33-50.
23. Austin PC, Ghali WA, Tu JV. A comparison of several regression
models for analysing cost of CABG surgery. Stat Med. 2003;22:
2799-2815.
24. Earle CC, Landrum MB, Souza JM, et al. Aggressiveness of cancer
care near the end of life: is it a quality-of-care issue? J Clin Oncol.
2008;26:3860-3866.
25. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for
patients with metastatic non–small-cell lung cancer. N Engl J Med.
2010;363:733-742.
26. Zimmermann C, Swami N, Krzyzanowska M, et al. Early palliative
care for patients with advanced cancer: a cluster-randomised con-
trolled trial. Lancet. 2014;383:1721-1730.
27. Walker H, Anderson M, Farahati F, et al. Resource use and costs of
end-of-life/palliative care: Ontario adult cancer patients dying during
2002 and 2003. J Palliat Care. 2011;27:79-88.
Cancer September 15, 2015
Cost of Aggressive End-of-Life Care/Cheung et al
28. Morishima T, Lee J, Otsubo T, et al. Association of healthcare
expenditures with aggressive versus palliative care for cancer patients
at the end of life: a cross-sectional study using claims data in Japan.
Int J Qual Health Care. 2014;26:79-86.
29. Longo CJ, Deber R, Fitch M, et al. An examination of cancer
patients’ monthly ‘out-of-pocket’ costs in Ontario, Canada. Eur J
Cancer Care (Engl). 2007;16:500-507.
30. Dumont S, Jacobs P, Turcotte V, et al. The trajectory of palliative
care costs over the last 5 months of life: a Canadian longitudinal
study. Palliat Med. 2010;24:630-640.
31. Jacobson M, O’Malley AJ, Earle CC, et al. Does reimbursement
influence chemotherapy treatment for cancer patients? Health Aff
(Millwood). 2006;25:437-443.
32. Matsuyama R, Reddy S, Smith TJ. Why do patients choose chemo-
therapy near the end of life? A review of the perspective of those fac-
ing death from cancer. J Clin Oncol. 2006;24:3490-3496.
3315
 JNCI J Natl Cancer Inst (2015) 107(3): dju424

doi:10.1093/jnci/dju424
First published online January 20, 2015
Article

article
Palliative Care and the Aggressiveness of End-of-Life
Care in Patients With Advanced Pancreatic Cancer
Raymond W. Jang, Monika K. Krzyzanowska, Camilla Zimmermann,

Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016

Nathan Taback, Shabbir M. H. Alibhai
Affiliation of authors:Division of Medical Oncology and Hematology (RWJ, MKK) and Department of Psychosocial Oncology and Palliative Care (CZ), Princess Margaret
Cancer Centre/University Health Network, Toronto, Canada (RWJ, MKK); Department of Medicine (RWJ, MKK, CZ, SMHA), Institute of Medical Sciences (MKK, CZ,
SMHA), Division of Biostatistics, Dalla Lana School of Public Health (NT), and Institute of Health Policy, Management, and Evaluation (MKK, SMHA), University of
Toronto, Toronto, Canada; Institute for Clinical Evaluative Sciences, Toronto, Canada (MKK).

Correspondence to: Raymond W. Jang, Princess Margaret Cancer Centre, 610 University Avenue, Rm 5–217, Toronto, ON M5G 2M9, Canada
(e-mail: raymond.jang@uhn.ca).

Abstract
Background: We examined the impact of palliative care (PC) on aggressiveness of end-of-life care for patients with
advanced pancreatic cancer. Measures of aggressive care included chemotherapy within 14 days of death; and at least one
intensive care unit (ICU) admission, more than one emergency department (ED) visit, and more than one hospitalization, all
within 30 days of death.
Methods: A retrospective population-based cohort study using administrative data was conducted in patients with
advanced pancreatic cancer from 2005 to 2010 in Ontario, Canada. Multivariable logistic regression was performed with
the above measures of aggressive care as the outcomes of interest and PC as the main exposure, adjusting for covariables.
Secondary analyses examined intensity of PC as the main exposure defined in two ways: 1) absolute number of PC visits
before the outcome of interest (0, 1, 2, 3+ visits) and 2) monthly rate of PC visits.
article

Results: The cohort included 5381 patients (median survival 75 days); 2816 (52.3%) had received a PC consultation. PC
consultation was associated with decreased use of chemotherapy near death (odds ratio [OR] = 0.34, 95% confidence
interval [CI] = 0.25 to 0.46); lower risk of ICU admission: OR = 0.12, 95% CI = 0.08 to 0.18; multiple ED visits: OR = 0.19, 95%
CI = 0.16 to 0.23; multiple hospitalizations near death: OR = 0.24, 95% CI = 0.19 to 0.31). A per-unit increase in the monthly
rate of PC visits was associated with lower odds of aggressive care for all four outcomes.
Conclusion: PC consultation and a higher intensity of PC were associated with less aggressive care near death in patients
with advanced pancreatic cancer.

Pancreatic cancer is a common and lethal type of cancer (1).
Surgery is the only curative treatment, but only around 20% of
patients are surgical candidates because most have advanced
disease at diagnosis. Even among those who have a poten-
tially curative surgical resection, the five-year survival is only
20% (2). Patients with pancreatic cancer also have a high bur-
den of symptoms and are frequent users of the emergency
department (ED). In a recent study, patients with pancreatic
cancer were second only to those with lung cancer in the fre-
quency of ED visits in both the last six months and the last two
weeks of life (3).
Given that the majority of patients with pancreatic cancer
die in a short period of time and with a high symptom burden,
how they die, their quality of life near death, and the quality

Received: April 18, 2014; Revised: October 13, 2014; Accepted: November 26, 2014
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

1 of 8

of care that they receive are important. For these reasons, pan-
creatic cancer is a good prototype cancer to study the quality
of care that patients receive near death. Aggressive care near
death for patients with incurable cancers is an important end-
of-life quality indicator (4,5), as aggressive care has been linked
with worse quality of life for the patient and worse bereavement
adjustment for their caregivers (6,7).
The most common indicators used to study aggressive
care near death include the administration of chemotherapy
near death, frequent ED visits near death, hospitalizations
near death (including death in a hospital), and intensive care
unit (ICU) admissions near death (4,5,8–13). However, quality
indicators are only useful if they are linkable to interventions
that can be shown to improve areas of suboptimal perfor-
mance. Because of its patient-centered approach and focus
on pain and symptom management (14), palliative care (PC)
may be an ideal intervention to reduce aggressiveness of care
near death.
Several randomized controlled trials have examined the
effect of different interventions on the use of health care
resources related to aggressiveness of care, such as num-
ber of ED visits (15–18), hospitalizations (15–22), ICU admis-
sions (15), advanced care planning (23–25), and death at home
(19,22,26–28) in cancer patients. It is difficult to draw defini-
tive conclusions from these studies, as the results were mixed
with respect to the effects of the intervention on health care
use (29,30), and only a few of those studies included a PC phy-
sician consultation as part of the intervention and a patient
population consisting of mostly cancer patients (17,20,27,31).
The best evidence comes from a single-center academic ran-
domized controlled trial that used early PC as an intervention
for a cohort of patients with metastatic non–small cell lung
cancer and found that patients randomly assigned to early PC
experienced better quality of life, better mood, and longer sur-
vival, while receiving less aggressive care near death (31). Some
retrospective evidence on the effects of PC on aggressive care
exists (11,32–35); however, no large scale, population-based
studies have examined this issue.
The purpose of this study was to evaluate the impact of PC,
including intensity of PC, on the aggressiveness of care near
death in patients with advanced pancreatic cancer. Our hypoth-
esis was that PC would be associated with decreased aggressive
care near death for these patients.
Methods
Study Design
We conducted a population-based, retrospective study of
an incident cohort of all patients who were diagnosed with
advanced pancreatic cancer in Ontario, Canada, between Jan 1,
2005 and Dec 31, 2010 and who subsequently died. Follow-up
data were available until March 2011. We included only patient
cases with a valid provincial health insurance number and
age 18 years or older at the time of diagnosis. Ethics approval
was obtained from the University Health Network and the
University of Toronto. Patient informed consent was waived, as
the data was deidentifed and anonymized before given to the
investigators.
Data Sources
We used the Ontario Cancer Registry (OCR) to identify index
cases of patients with a diagnosis of pancreatic cancer. The
OCR is a population-based cancer registry and is approximately
R. W. Jang et al.  |  2 of 8
95% complete (36). Index cases from the OCR were linked using
encrypted provincial health card numbers across different data
sources. The data sources included: 1) Ontario Health Insurance
Plan (OHIP) (37) claims database, which contains information on
billing claims for physicians’ services provided to Ontario resi-
dents, 2)  the Canadian Institute for Health Information (CIHI)
Discharge Abstract Database (DAD) (38), which contains diag-
nostic and procedure information on all discharges from acute
care facilities and information on same-day surgeries, 3)  the
CIHI National Ambulatory Care Reporting System (NACRS) (39),
which captures information on emergency department visits,
and 4) the Registered Persons Database (RPDB), which provides
basic sociodemographic information.
Cohort Selection
The OCR does not contain staging information for pancreatic
cancer. Therefore, in order to avoid the bias of including patients
Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016
with potentially curable disease (40), we created a proxy defi-
nition of advanced stage (ie, locally advanced and metastatic)
by restricting patients with pancreatic cancer using the follow-
ing criteria. Patients were excluded if they had: 1) a diagnosis of
another cancer (either before or after the diagnosis of pancreatic
cancer), as it would not be known which cancer contributed to
the death, 2)  major pancreatic cancer surgery with no chemo-
therapy, or 3)  major pancreatic cancer surgery with chemo-
therapy within 120 days of surgery. Excluding patients who have
had cancer-directed surgery to define a cohort of patients with
advanced cancer has been used in a previous study of end-of-
life care (12).
Study Objectives
Our main objective was to evaluate the impact of an inpatient
or outpatient PC consultation in patients with advanced pan-
creatic cancer on the proportions of: 1)  chemotherapy near
death (received the last dose of chemotherapy within 14  days
of death), 2)  multiple ED visits near death (more than one ED
visit within 30 days of death), 3) multiple hospitalizations near
death (more than one hospitalization within 30 days of death),
and 4)  ICU admission near death (at least one ICU admission
article
within 30 days of death).
A secondary objective was to study the impact of the inten-
sity of PC using the same four measures of aggressive care. The
intensity of PC was defined in two ways: 1) absolute number of
PC visits before the outcome of interest (0, 1, 2, 3+ visits) and
2)  monthly rate of PC visits defined as: number of PC visits ÷
(survival time in days ÷ 30). Patients who lived less than 14 days
were excluded, as their limited survival would have artificially
inflated the monthly rate (eg, if a person has one PC visit but
only lived for one day, the monthly rate would be 1/[1/30] = 30
visits). The PC visits were counted from the day of diagnosis
until each outcome of interest.
The four outcome measures of aggressive end-of-life care
were based on previous studies (5,41,42) and used the same
codes as a previous Ontario study (41). PC was defined using
OHIP physician billing data. Data sources and definitions of
important variables are provided in Supplementary Table 1.
Statistical Analysis
We calculated the frequency of each of the measures of aggres-
sive care for all patients and conducted subset analyses for
those who received or did not receive PC. The denominator
 3 of 8  |  JNCI J Natl Cancer Inst, 2015, Vol.107, No.3
included all patients who died with a diagnosis of an advanced
pancreatic cancer. Chi-squared (χ2) tests were performed as
a first test to determine differences in aggressiveness of care
among patients who did and did not receive palliative care. All
statistical tests were two-sided and a P value of less than .05 was
considered statistically significant.
Multivariable logistic regression analyses were used to pre-
dict the likelihood of experiencing each of the outcomes of
aggressive care, with PC as the main exposure and the following
variables as covariables: age, sex, rural residence, health region,
year of diagnosis, income quintile, primary care physician
involvement, medical oncologist involvement (for outcomes
other than chemotherapy near death), chemotherapy (for out-
comes other than chemotherapy near death), and Charlson-Deyo
Comorbidity Index Score (0, 1, 2+) (43). The index was obtained
from hospitalizations and same-day surgery admissions only.
Similar analyses were performed for the secondary objective,
using the two definitions of intensity of PC in separate models.
All analyses were performed with the R statistical software
(http://www.r-project.org; version 2.14.0).
Sensitivity Analysis
We conducted several types of sensitivity analyses. First, we lim-
ited the cohort to those who had received some palliative chemo-
therapy. These patients are different in that they are likely to have
better performance status and have been seen by a medical oncol-
ogist. Second, we excluded patients diagnosed in 2010, as these
will have had a shorter follow-up. Third, we included survival
duration as a covariable in the logistic regression, as patients who
lived for a longer time have more exposure time to receive PC, and
differing survival times may have different effects on aggressive
care near death. The multivariable logistic regression analyses
were repeated by adding survival duration as a continuous covari-
able (in days) and as a categorical covariable (short duration: fewer
than 90  days; medium duration: 90 or more days and less than
180 days; long duration: 180 or more days) in separate models.
Fourth, we performed a hierarchical random effects model
(44) with the patient’s specific medical oncologist as the clus-
tering variable (level 1), while keeping all other covariables
unchanged (level 2, patient-level), except for removal of the
article
binary covariable of having seen a medical oncologist from level
2. This was done using each physician’s uniquely anonymized
identifier from the OHIP database. Medical oncologists were
chosen as the clustering variable, as some medical oncologists
may be more likely to refer to palliative care, while recommend-
ing less aggressive treatment, creating a possible bias for our
study. We considered both palliative care and primary care
physicians as alternative clustering variables. Palliative care
physicians were felt to be an inappropriate clustering variable
since they constituted the main exposure. We felt that family
physicians usually are not the ones involved in decision-mak-
ing such as chemotherapy near death and ICU admission near
death. Some family physicians are also the local palliative care
doctors and do home visits—when these doctors billed the pal-
liative care billing codes, the patient was considered as having
received palliative care. Analysis of within-cluster variance was
also performed. The random effects model was chosen to be a
sensitivity analysis as opposed to the main analysis, as approxi-
mately 40% of the patients in our cohort had not seen a medical
oncologist.
We also performed a propensity score analysis based on the
receipt of PC, as those receiving PC may be different in ways
that cannot be measured than those who did not. We created
a propensity score for the likelihood of receiving PC, ranked all
the patients, created quintiles based on this score, and included
the scores as an independent variable in our four main regres-
sion analyses.
Lastly, we examined the effect of the medical oncologist on
each of our outcomes by using a two-level hierarchical random
effects model. The specific methodology and results are outlined
in the supplemental file.
Results
Cohort Description and Patient Demographics
In our initial cohort, 7928 patients were diagnosed with pancre-
atic cancer in Ontario. Of these, 670 were excluded because they
had a diagnosis of a second cancer, leaving 7258 patients. Of
the 7258 patients, 1182 were excluded because they underwent
pancreatic cancer surgery (533 without adjuvant chemotherapy,
Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016
and 649 with adjuvant chemotherapy). The remaining 6076
patients were considered to have advanced disease. Of these,
5381 (88.6%) had died by March 2011 and comprised our cohort.
Table 1 shows the patient characteristics. Patients age 80 years
and older made up 29.1% of the final cohort. Males and females
were equally represented, 14.6% of patients lived in rural areas,
81.5% of patients had a primary care physician, and 63.1% of
patients had seen a medical oncologist.
Palliative Care Visit Characteristics and
Survival Times
Table  1 also displays characteristics of PC visits and survival
times. Approximately half of the patients (2816 of 5381 or 52.3%)
had a PC consultation: 9.5% had one visit, 10.3% had two visits,
and 32.5% had three or more visits. For patients who had seen a
PC physician, the median and mean numbers of PC visits were
3 and 5.7, respectively (interquartile range [IQR]  =  2 to 7). The
median and mean survival times of patients with advanced pan-
creatic cancer were 75 and 147 days, respectively. Of those who
received some palliative chemotherapy, the median and mean
survival times were 199 and 264 days, respectively. Patients who
received PC also had longer survival times (eg, median survival
of 100 days vs 51 days) compared with those who did not.
Frequency of Aggressive Care Near Death and
Association With PC
Of the 5381 patients, 4.1% received chemotherapy near death,
4.3% had ICU admissions near death, 18.5% had multiple ED
visits near death, and 8.3% had multiple hospitalizations near
death; 74.1% of patients did not experience any of the four meas-
ures of aggressive care near death (Table 2). Table 2 also shows
the unadjusted frequency of each outcome for all patients and
by receipt of PC. Patients who had received a PC consultation
had lower unadjusted frequencies of aggressive care than those
who did not: 2.6% vs 5.6% (χ2 = 29.9, P < .001) for chemotherapy
near death, 1.1% vs 7.8% (χ2 = 142.3, P < .001) for ICU near death,
7.4% vs 28.5% (χ2  =  393.5, P < .001) for multiple ED visits, and
3.8% vs 12.8% (χ2 = 141.8, P < .001) for multiple hospitalizations.
Figure 1 illustrates the relationship between aggressive care and
number of PC visits.
Table 3 shows the results of the multivariable logistic regres-
sion analyses. PC consultation was associated with lower odds
of receiving any of the four measures of aggressive care near
 R. W. Jang et al.  |  4 of 8

death: chemotherapy near death (OR  =  0.34, 95% CI  =  0.25 to hospitalizations near death (OR  =  0.24, 95% CI  =  0.19 to 0.31).
0.46); ICU admission (OR  =  0.12, 95% CI  =  0.08 to 0.18), mul- More frequent PC visits were associated with fewer instances of
tiple ED visits (OR  =  0.19, 95% CI  =  0.16 to 0.23), and multiple aggressive care, as seen in Table 4.

Table 1.  Pancreatic cohort demographics (n = 5381)

Other Variables Associated With Aggressive Care
Characteristic No. (%)
As seen in Table  3, the oldest (80+ y) patients received less
Age, y aggressive care than the youngest patients (<50 y). Male patients
 <50 246 (4.6%) had higher proportions of aggressive care than female patients,
 50–59 736 (13.7%) although this was only statistically significant for two measures:
 60–69 1205 (22.4%)
ICU near death (OR  =  1.46, 95% CI  =  1.09 to 1.94, P  =  .01) and
 70–79 1629 (30.3%)
multiple hospitalizations near death (OR  =  1.30, 95% CI  =  1.06
 80+ 1565 (29.1%)
to 1.60, P  =  .01). Patients in rural areas had higher proportions
Sex (female) 2753 (51.2%)
of multiple ED visits near death, but lower proportions of ICU
Rural 783 (14.6%)
admission near death. Patients in higher income quintiles had
Income quintile
higher proportions of chemotherapy near death, and this was
  Missing value 29 (0.5%)
 1 1128 (21.0%) statistically significant when comparing the highest income
 2 1130 (21.0%) quintile with the lowest. Patients who saw a medical oncolo-

Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016

 3 1054 (19.6%) gist had lower proportions of ICU admission near death. Other
 4 1081 (20.1%) covariables such as year of diagnosis and having a primary care
 5 959 (17.8%) physician were not associated with appreciable differences in
Year of diagnosis the frequency of aggressive care (Table 3).
 2005 938 (17.4%)
 2006 902 (16.8%)
Sensitivity Analyses
 2007 1000 (18.6%)
 2008 978 (18.2%) PC was still associated with lower odds of receiving aggressive
 2009 901 (16.7%) care near death in all of the sensitivity analyses (Table  5). For
 2010 662 (12.3%) example, for the outcome of chemotherapy near death, the odds
Primary care physician 4383 (81.5%)
ratio from the main model was 0.34, while the odds ratios from
Medical oncologist 3393 (63.1%)
the other models ranged from 0.22 to 0.35. Specifically, the hier-
Number of medical oncologists in 281
archical random effects model using specific medical oncolo-
our study
gists as the clustering variable produced similar results. The
Charlson Index Score
effects of medical oncologists on each outcome are shown in
  No Charlson score* 132 (2.5%)
Supplementary Table 2 (available online). Having a specific medi-
 0 3700 (68.8%)
 1 982 (18.2%) cal oncologist had an impact on the likelihood of chemotherapy
  2 or more 567 (10.5%) near death and ICU near death, but not multiple ED visits and
PC consult 2816 (52.3%) multiple hospitalizations near death.
Number of PC visits (on distinct days)
 0 2565 (47.7%) Discussion
 1 513 (9.5%)
 2 555 (10.3%) Aggressive care near death has been found to lead to worse qual-

article
 3+ 1748 (32.5%) ity of life for the patient and worse bereavement adjustment for
Survival (all patients) their caregivers (6,7). Despite recent evidence on the benefits
  Median (mean) 75 (147) days of PC (31,45), few studies have examined the impact of PC on
  Interquartile Range 27–188 days aggressiveness of care near death from a population perspec-
Survival (patients who received palliative chemotherapy) tive. In our cohort of patients with advanced pancreatic cancer,
  Median (mean) 199 (264) days a PC consultation was associated with less aggressive care near
Survival by receipt of PC death. This finding was consistent across all four measures of
  Median (mean) aggressive care near death (ie, chemotherapy, ICU admissions,
PC 100 (172) days multiple ED visits, and hospitalizations). The magnitude of ben-
  No PC 51 (120) days
efit was also large, as all the odds ratios ranged from 0.12 to 0.34
* These patients do not have a calculable Charlson score, because they were not
on multivariable analyses. Furthermore, we found that more
hospitalized or did not have same-day surgery during the two years prior to intensive PC follow-up was associated with fewer instances of
diagnosis to last follow-up. PC = palliative care. aggressive care near death.

Table 2.  Unadjusted frequency of each outcome for all patients and by receipt of PC (advanced cohort; n = 5381)*

Outcome All patients, No. (%) PC, No. (%) No PC, No. (%)

Chemotherapy near death 218/5381 (4.1) 73/2790 (2.6) 145/ 2591 (5.6)
ICU near death 234/5381 (4.3) 31/2775 (1.1) 203/2606 (7.8)
Multiple ED visits near death 993/5381 (18.5) 190/2560 (7.4) 803/2821 (28.5)
Multiple hospitalizations near death 447/5381 (8.3) 103/2697 (3.8) 344/2684 (12.8)

* To be considered as having received palliative care (PC), the PC consultation had to occur before the outcome of interest. ED = emergency department; ICU = inten-
sive care unit; PC = palliative care.
 5 of 8  |  JNCI J Natl Cancer Inst, 2015, Vol.107, No.3

Multiple hospitalizations near death

0.24 (0.19 to 0.31)

0.41 (0.26 to 0.66)

1.06 (0.83 to 1.35)
1.02 (0.79 to 1.32)
0.92 (0.68 to 1.23)

0.78 (0.56 to 1.11)

0.92 (0.72 to 1.18)
0.82 (0.57 to 1.17)
1.12 (0.78 to 1.62)
1.30 (1.06 to 1.60)
OR (95% CI)

Referent

Referent
Figure  1.  Frequency of aggressive care by number of palliative care visits.
ED = emergency department; ICU = intensive care unit.

* Models were also adjusted for health region, but this variable is not shown for ease of presentation. ED = emergency department; ICU = intensive care unit; OR = odds ratio.
Multiple ED visits near death
Multiple sensitivity analyses, including using only the subco-

Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016

hort of patients who received palliative chemotherapy, adjust-

0.44 (0.30 to 0.65)

1.01 (0.85 to 1.20)
0.82 (0.67 to 0.99)

1.03 (0.89 to 1.20)

0.96 (0.75 to 1.22)

1.30 (0.99 to 1.71)
1.17 (0.92 to 1.49)
1.30 (0.99 to 1.71)
0.19 (0.16 to 0.23)

1.37 (1.11 to 1.69)

ing for survival duration, and alternative modelling approaches

OR (95% CI)
(hierarchical modelling and propensity score modelling) did not

Referent

Referent
appreciably change our main findings. Specifically, the hierar-
chical random effects model with the medical oncologist as
the clustering variable provides evidence that the results are
not simply because of certain medical oncologists recommend-
ing both a palliative care consultation and less aggressive care
near death.
There are several elements of a palliative care consultation
that could explain its association with decreased aggressiveness
of care. One likely element is that of simply having an end-of-
life conversation, as this has been shown to decrease aggres-
sive care near death (46). A recent qualitative study associated
0.12 (0.08 to 0.18)

0.12 (0.06 to 0.23)

0.50 (0.36 to 0.69)
0.77 (0.53 to 1.13)
0.61 (0.38 to 0.98)

1.27 (0.75 to 2.15)

1.36 (0.94 to 1.97)
1.29 (0.85 to 1.96)
1.27 (0.75 to 2.15)
1.46 (1.09 to 1.94)
ICU near death

with the trial by Temel and colleagues (31) described seven key
OR (95% CI)

Referent

Referent
elements to a PC consultation: relationship and rapport build-
ing, symptom assessment, addressing coping issues, estab-
lishing illness understanding, discussing cancer treatments,
Table 3.  Logistic regression analyses predicting aggressive care near death (n = 5381)*

end-of-life planning, and engagement of family members (47).

The emphasis on specific elements during palliative care visits
changed over the illness trajectory, with initial visits focusing on
relationship building and illness understanding, and later visits
focusing on resuscitation preferences and hospice discussions
article

(47). These qualitative data, together with our quantitative find-

ings that more PC visits were associated with fewer instances of
Chemotherapy near death

aggressive care, suggest that there might be a minimum number

0.06 (0.03 to 0.13)

0.76 (0.48 to 1.20)

1.24 (0.93 to 1.66)

1.12 (0.78 to 1.61)

0.74 (0.43 to 1.28)
1.51 (0.88 to 2.61)
1.90 (1.19 to 3.02)
0.34 (0.25 to 0.46)

of visits that are needed to benefit fully from PC. Although it is

OR (95% CI)

not clear how many visits are necessary, in the study by Temel
Referent

Referent

and colleagues (31) the average number of visits in the PC group

N/A
N/A

was four, whereas in ours, the median was three visits.

The intensity of PC is linked to its timing. Certainly early PC
likely leads to more PC visits. However, it does not necessar-
ily follow that a referral has to be made within a certain time
interval from diagnosis for the patient to have a sufficiently long
time with PC to benefit. One possible referral system would be
based on prognosis (based on clinician judgment and/or vali-
dated prognostic tools) (48,49) and/or symptoms (based on the
Edmonton Symptom Assessment Score) (50).
Primary care physician

Other findings of our study were that older patients and

women were less likely to receive aggressive care near death,
  Highest quintile
  Lowest quintile
Medical oncology

Year of diagnosis

which is consistent with previous studies (11,41,51). It is not sur-

Increasing age, y

Income quintile

Charlson index
Chemotherapy
Palliative care

prising that older patients received less aggressive care, although

it is unclear why care for women was less aggressive. As has been
  Ref. <50
Predictor

Male sex
Rurality

seen in a previous study (11), rural patients were less likely to be

 80+

admitted to the ICU but more likely to have multiple ED visits near
death. These findings likely reflect the lack of certain health care
 R. W. Jang et al.  |  6 of 8

Table 4.  Logistic regression analyses using intensity of PC

Chemotherapy near death ICU near death Multiple ED visits near death Multiple hospitalizations near death

Predictor OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)

Number of PC visits
  Ref. 0 Referent Referent Referent Referent
 1 0.73 (0.47 to 1.16) 0.25 (0.13 to 0.46) 0.36 (0.27 to 0.48) 0.47 (0.33 to 0.69)
 2 0.26 (0.14 to 0.51) 0.13 (0.06 to 0.29) 0.29 (0.22 to 0.40) 0.44 (0.30 to 0.64)
 3+ 0.26 (0.17 to 0.38) 0.09 (0.05 to 0.16) 0.11 (0.09 to 0.15) 0.12 (0.09 to 0.18)
Monthly rate of PC
visits, visits/month
  Ref. 0 Referent Referent Referent Referent
 1 0.69 (0.58 to 0.81) 0.31 (0.21 to 0.46) 0.31 (0.26 to 0.38) 0.31 (0.24 to 0.40)

* Patients who lived less than 14 days were excluded, as their limited survival would have artificially inflated the monthly rate (eg, if a person has one palliative care visit
but only lived for one day, the monthly rate would be 1/[1/30] = 30 visits). ED = emergency department; ICU = intensive care unit; OR = odds ratio; PC = palliative care.

Table 5.  Sensitivity analysis: comparison of the effect of palliative care using different models

Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016

Logistic regression Logistic regression
Logistic regres- model (palliative Logistic regression model with survival
sion model (main chemotherapy model with the duration as a co- Hierarchical ran- Propensity score
analysis) patients only) 2010 year removed variable* dom effects model† model

Adjusted odds ratio of the effect of palliative care on the specified outcome (95% CI)

Outcome n = 5381 n = 1664 n = 4719 n = 5381 n = 3393 n = 5381

Chemotherapy
near death
  Ref. No PC Referent Referent Referent Referent Referent Referent
 PC 0.34 (0.25 to 0.46) 0.22 (0.16 to 0.30) 0.30 (0.22 to 0.43) 0.34 (0.25 to 0.47) 0.29 (0.21 to 0.39) 0.35 (0.26 to 0.48)
ICU near death
  Ref. No PC Referent Referent Referent Referent Referent Referent
 PC 0.12 (0.08 to 0.18) 0.21 (0.11 to 0.39) 0.13 (0.08 to 0.19) 0.12 (0.08 to 0.18) 0.15 (0.10 to 0.25) 0.13 (0.09 to 0.19)
Multiple ED visits
near death
  Ref. No PC Referent Referent Referent Referent Referent Referent
 PC 0.19 (0.16 to 0.23) 0.36 (0.28 to 0.48) 0.19 (0.16 to 0.23) 0.20 (0.17 to 0.24) 0.22 (0.18 to 0.27) 0.20 (0.17 to 0.24)
Multiple hospitalizations
near death
  Ref. No PC Referent Referent Referent Referent Referent Referent
 PC 0.24 (0.19 to 0.31) 0.24 (0.19 to 0.31) 0.24 (0.19 to 0.31) 0.25 (0.20 to 0.32) 0.29 (0.22 to 0.38) 0.25 (0.20 to 0.32)

article
*  The survival duration covariable was a categorical variable: short (<90 days), medium (≥90 days and < 180 days), and long (≥180 days).
† For the random effects model, the n of 3393 is smaller than the overall n of 5381 because only approximately 60% of patients saw a medical oncologist, and
the clustering is around individual medical oncologists. The binary covariable of medical oncologist (TRUE/FALSE) was excluded. ED = emergency department;
ICU = intensive care unit; PC = palliative care.

resources available in rural areas. At a more general level, these
findings suggest that the different measures of aggressive care
measure different aspects of such care, and that, at least for some
measures of aggressiveness, the input variables go beyond that of
a physician and the patient and family.
Our study has several limitations. As a way to overcome the
biases and limitations inherent with observational studies, we
performed multiple sensitivity analyses. However, we cannot
correct for sources of bias that cannot be measured, and ulti-
mately we cannot determine causality (ie, that a PC consulta-
tion itself is causing the decrease in aggressive care near death).
It is conceivable that the patients and families themselves are
the main determinants of receipt of aggressive care and that
PC is simply a “marker” of their shift in thinking from focusing
on cure to maximizing quality of life. However, even in these
instances, patients benefit from physician (oncologist or pal-
liative care) guidance in order to make informed decisions and
make appropriate plans. Additionally, our effectiveness findings
are similar to that of the pivotal efficacy trial by Temel et al. (31).
Other limitations of our study include the fact that admin-
istrative databases are neither 100% complete or accurate, the
unavailability of variables such as patient performance status
and symptom burden, and the lack of validation of certain billing
codes. Case ascertainment bias may exist, as not all patients would
have had pathologic confirmation of the diagnosis. Selection bias
exists in how we defined the final cohort. In our goal of exclud-
ing patients with potentially curable disease, some patients who
ultimately had incurable disease might have been missed (eg,
patients who received pancreatic surgery and recurred early).
However, the magnitude of the bias is likely small. Lastly, although
we controlled for survival duration in our sensitivity analysis, the
persistent survival difference seen in those who received pallia-
tive care and those who did not, may indicate residual confound-
ing that was not corrected even with the sensitivity analysis.
Despite these limitations, our study adds important knowl-
edge to what is currently known about the effects of PC,
using a population-based design of more than 5000 patients,
a cancer other than lung cancer, and multiple measures of
 7 of 8  |  JNCI J Natl Cancer Inst, 2015, Vol.107, No.3
aggressiveness. The results suggest that at a population level,
PC is associated with decreased aggressive care near death.
Our work provides an effectiveness analysis of PC in routine
clinical practice that clinical trials cannot provide, thereby
extending the important work of two recent clinical trials
(31,45). Given the evidence that aggressive care has been asso-
ciated with worse quality of life for patients and their families,
is often incongruent with patient wishes (52), and is increas-
ing over time (41,51), it is our hope that our study will add an
important piece to the growing literature on the benefits of
PC and provide an impetus to increase the availability of PC
worldwide.
Funding
This work was supported by a 2011 Young Investigator Award
from the Conquer Cancer Foundation of the American Society
of Clinical Oncology, an award from the University of Toronto
Medical Oncology Postgraduate Ontario Student Opportunity
Trust Funds (OSOTF), and the Division of Medical Oncology and
Hematology, Princess Margaret Cancer Centre, University Health
Network to RWJ.
This study was supported through provision of data by the
Institute for Clinical Evaluative Sciences (ICES) and Cancer
Care Ontario (CCO) and through funding support to ICES from
an annual grant by the Ministry of Health and Long-Term Care
(MOHLTC).
Notes
This paper was presented in part at the European Society for
Medical Oncology (ESMO) Annual Congress, Vienna, October
1, 2012, and the American Society of Clinical Oncology (ASCO)
Annual Meeting, Chicago, June 1, 2013.
The study funders had no role in the design of the study, the
collection, analysis, or interpretation of the data, the writing of
the manuscript, nor the decision to submit the manuscript for
publication. No endorsement by ICES, CCO, or the MOHLTC is
intended or should be inferred.
References
article
1. Canadian Cancer Statistics 2013: Canadian Cancer Society;
2013.
2. Mayo SC, Nathan H, Cameron JL, et  al. Conditional survival
in patients with pancreatic ductal adenocarcinoma resected
with curative intent. Cancer. 2012;118(10):2674–2681.
3. Barbera L, Taylor C, Dudgeon D. Why do patients with cancer
visit the emergency department near the end of life? CMAJ.
2010;182(6):563–568.
4. Earle CC, Park ER, Lai B, et al. Identifying potential indicators
of the quality of end-of-life cancer care from administrative
data. J Clin Oncol. 2003;21(6):1133–1138.
5. Grunfeld E, Urquhart R, Mykhalovskiy E, et  al. Toward pop-
ulation-based indicators of quality end-of-life care: testing
stakeholder agreement. Cancer. 2008;112(10):2301–2308.
6. Wright AA, Zhang B, Ray A, et al. Associations between end-
of-life discussions, patient mental health, medical care
near death, and caregiver bereavement adjustment. JAMA.
2008;300(14):1665–1673.
7. Wright AA, Keating NL, Balboni TA, et al. Place of death: cor-
relations with quality of life of patients with cancer and pre-
dictors of bereaved caregivers’ mental health. J Clin Oncol.
2010;28(29):4457–4464.
8. Earle CC, Neville BA, Landrum MB, et  al. Evaluating claims-
based indicators of the intensity of end-of-life cancer care.
Int J Qual Health Care. 2005;17(6):505–509.
9. Keating NL, Landrum MB, Lamont EB, et  al. End-of-life care
for older cancer patients in the Veterans Health Administra-
tion versus the private sector. Cancer. 2010;116(15):3732–3739.
10. Earle CC, Landrum MB, Souza JM, et  al. Aggressiveness of
cancer care near the end of life: is it a quality-of-care issue? J
Clin Oncol. 2008;26(23):3860–3866.
11. Barbera L, Paszat L, Chartier C. Indicators of poor quality end-
of-life cancer care in Ontario. J Palliat Care. 2006;22(1):12–17.
12. Warren JL, Barbera L, Bremner KE, et  al. End-of-life care for
lung cancer patients in the United States and Ontario. J Natl
Cancer Inst. 2011;103(11):853–862.
13. Krzyzanowska MK, Barbera L, Elit L, et al. Identifying popula-
tion-level indicators to measure the quality of cancer care for
women. Int J Qual Health Care. 2011;23(5):554–564.
14. WHO Definition of Palliative Care. http://www.who.int/cancer/
palliative/definition/en/. Accessed December 16, 2014.
Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016
15. Bakitas M, Lyons KD, Hegel MT, et al. Effects of a palliative care
intervention on clinical outcomes in patients with advanced
cancer: the Project ENABLE II randomized controlled trial.
JAMA. 2009;302(7):741–749.
16. Brumley R, Enguidanos S, Jamison P, et al. Increased satisfac-
tion with care and lower costs: results of a randomized trial of
in-home palliative care. J Am Geriatr Soc. 2007;55(7):993–1000.
17. Hughes SL, Cummings J, Weaver F, et al. A randomized trial of
the cost effectiveness of VA hospital-based home care for the
terminally ill. Health Serv Res. 1992;26(6):801–817.
18. Zimmer JG, Groth-Juncker A, McCusker J. A randomized con-
trolled study of a home health care team. Am J Public Health.
1985;75(2):134–141.
19. Moore S, Corner J, Haviland J, et al. Nurse led follow up and
conventional medical follow up in management of patients
with lung cancer: randomised trial. BMJ. 2002;325(7373):1145.
20. Hanks GW, Robbins M, Sharp D, et  al. The imPaCT study: a
randomised controlled trial to evaluate a hospital palliative
care team. Br J Cancer. 2002;87(7):733–739.
21. McCorkle R, Benoliel JQ, Donaldson G, et  al. A randomized
clinical trial of home nursing care for lung cancer patients.
Cancer. 1989;64(6):1375–1382.
22. Kane RL, Wales J, Bernstein L, et al. A randomised controlled
trial of hospice care. Lancet. 1984;1(8382):890–894.
23. Gade G, Venohr I, Conner D, et al. Impact of an inpatient pal-
liative care team: a randomized control trial. J Palliat Med.
2008;11(2):180–190.
24. Rabow MW, Dibble SL, Pantilat SZ, et al. The comprehensive
care team: a controlled trial of outpatient palliative medicine
consultation. Arch Intern Med. 2004;164(1):83–91.
25. A controlled trial to improve care for seriously ill hospitalized
patients. The study to understand prognoses and preferences
for outcomes and risks of treatments (SUPPORT). The SUP-
PORT Principal Investigators. JAMA. 1995;274(20):1591–1598.
26. Rummans TA, Clark MM, Sloan JA, et al. Impacting quality of
life for patients with advanced cancer with a structured mul-
tidisciplinary intervention: a randomized controlled trial. J
Clin Oncol. 2006;24(4):635–642.
27. Jordhøy MS, Fayers P, Saltnes T, et al. A palliative-care inter-
vention and death at home: a cluster randomised trial. The
Lancet. 2000;356(9233):888–893.
28. Grande GE, Todd CJ, Barclay SI, et al. Does hospital at home
for palliative care facilitate death at home? Randomised con-
trolled trial. BMJ. 1999;319(7223):1472–1475.
29. Zimmermann C, Riechelmann R, Krzyzanowska M, et  al.
Effectiveness of specialized palliative care: a systematic
review. JAMA. 2008;299(14):1698–1709.

30. Smith TJ, Temin S, Alesi ER, et al. American Society of Clini-
cal Oncology provisional clinical opinion: the integration
of palliative care into standard oncology care. J Clin Oncol.
2012;30(8):880–887.
31. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care
for patients with metastatic non-small-cell lung cancer. N
Engl J Med. 2010;363(8):733–742.
32. Back AL, Li YF, Sales AE. Impact of palliative care case
management on resource use by patients dying of can-
cer at a Veterans Affairs medical center. J Palliat Med.
2005;8(1):26–35.
33. Seow H, Piet L, Kenworthy CM, et al. Evaluating a palliative
care case management program for cancer patients: the
Omega Life Program. J Palliat Med. 2008;11(10):1314–1318.
34. Penrod JD, Deb P, Luhrs C, et al. Cost and utilization outcomes
of patients receiving hospital-based palliative care consulta-
tion. J Palliat Med. 2006;9(4):855–860.
35. Tse DM, Chan KS, Lam WM, et  al. The impact of palliative
care on cancer deaths in Hong Kong: a retrospective study of
494 cancer deaths. Palliat Med. 2007;21(5):425–433.
36. Robles SC, Marrett LD, Clarke EA, et  al. An application of
capture-recapture methods to the estimation of complete-
ness of cancer registration. J Clin Epidemiol. 1988;41(5):495–
501.
37. Ontario Health Insurance Plan (OHIP). http://www.health.gov.
on.ca/en/public/programs/ohip/. Accessed December 16,
2014.
38. CIHI - DAD Metadata. http://www.cihi.ca/cihi-ext-portal/inter-
net/en/document/types+of+care/hospital+care/acute+care/
dad_metadata. Accessed December 16, 2014.
39. CIHI - National Ambulatory Care Reporting System (NACRS).
www.cihi.ca/nacrs. Accessed December 16, 2014.
40. Bach PB, Schrag D, Begg CB. Resurrecting treatment histories
of dead patients: a study design that should be laid to rest.
JAMA. 2004;292(22):2765–2770.
41. Ho TH, Barbera L, Saskin R, et al. Trends in the aggressiveness
of end-of-life cancer care in the universal health care system
of Ontario, Canada. J Clin Oncol. 2011;29(12):1587–1591.
R. W. Jang et al.  |  8 of 8
42. Barbera L, Elit L, Krzyzanowska M, et  al. End of life care
for women with gynecologic cancers. Gynecol Oncol.
2010;118(2):196–201.
43. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbid-
ity index for use with ICD-9-CM administrative databases. J
Clin Epidemiol. 1992;45(6):613–619.
44. Gelman A HJ. Data Analysis Using Regression and Multilevel
Regression and Multilevel/Hierarchical Models. New York: Cam-
bridge University Press; 2007.
45. Zimmermann C, Swami N, Krzyzanowska M, et  al. Early
palliative care for patients with advanced cancer: a cluster-
randomised controlled trial. Lancet. 2014;383(9930):1721–
1730.
46. Mack JW, Cronin A, Keating NL, et  al. Associations between
end-of-life discussion characteristics and care received
near death: a prospective cohort study. J Clin Oncol.
2012;30(35):4387–4395.
47. Yoong J, Park ER, Greer JA, et  al. Early palliative care in
advanced lung cancer: a qualitative study. JAMA Intern Med.
Downloaded from http://jnci.oxfordjournals.org/ at University of Adelaide on April 4, 2016
2013;173(4):283–290.
48. Maltoni M, Caraceni A, Brunelli C, et  al. Prognostic factors in
advanced cancer patients: evidence-based clinical recommen-
dations--a study by the Steering Committee of the European
Association for Palliative Care. J Clin Oncol. 2005;23(25):6240–6248.
49. Chow E, Abdolell M, Panzarella T, et  al. Predictive model
for survival in patients with advanced cancer. J Clin Oncol.
2008;26(36):5863–5869.
50. Bruera E, Kuehn N, Miller MJ, et  al. The Edmonton Symp-
tom Assessment System (ESAS): a simple method for
the assessment of palliative care patients. J Palliat Care.
1991;7(2):6–9.
51. Earle CC, Neville BA, Landrum MB, et al. Trends in the aggres-
siveness of cancer care near the end of life. J Clin Oncol.
2004;22(2):315–321.
52. Mack JW, Weeks JC, Wright AA, et al. End-of-life discussions,
goal attainment, and distress at the end of life: predictors
and outcomes of receipt of care consistent with preferences.
J Clin Oncol. 2010;28(7):1203–1208.
article